CHAPTER

9 Vector Calculus

In Chapter 7 we studied the

properties of vectors in 2- and CHAPTER CONTENTS
3-space. In this chapter we will
combine vector concepts with
those from differential and 9.1 Vector Functions
integral calculus. 9.2 Motion on a Curve
9.3 Curvature and Components of Acceleration
9.4 Partial Derivatives
9.5 Directional Derivative
9.6 Tangent Planes and Normal Lines
9.7 Curl and Divergence
9.8 Line Integrals
9.9 Independence of the Path
9.10 Double Integrals
9.11 Double Integrals in Polar Coordinates
9.12 Green’s Theorem
9.13 Surface Integrals
9.14 Stokes’ Theorem
9.15 Triple Integrals
9.16 Divergence Theorem
9.17 Change of Variables in Multiple Integrals
Chapter 9 in Review
 9.1 Vector Functions
y
(x(t0), y(t0)) INTRODUCTION Recall that a curve C in the xy-plane is simply a set of ordered pairs
(x, y). We say that C is a parametric curve if the x- and y-coordinates of a point on the curve are
defined by a pair of functions x  f (t), y  g(t) that are continuous on some interval a  t  b.
C The notion of a parametric curve extends to 3-space as well. A parametric curve in space, or
space curve, is a set of ordered triples (x, y, z) where

x  f (t), y  g(t), z  h(t), (1)

r(t0) = 〈x(t0), y(t0)〉
x
are continuous on an interval defined by a  t  b. In this section we combine the concepts of
(a) 2-space
parametric curves with vectors.
z
Vector-Valued Functions It is often convenient in science and engineering to introduce
r(t0) = 〈x(t0), y(t0), z(t0)〉 C
a vector r whose components are functions of a parameter t. We say that
(x(t0), y(t0), z(t0))
r(t)   f (t), g(t)  f (t) i  g(t) j

y and r(t)   f (t), g(t), h(t)  f (t) i  g(t) j  h(t) k,

are vector-valued functions or simply vector functions. As shown in FIGURE 9.1.1, for a given
x value of the parameter, say t0, the vector r(t0) is the position vector of a point P on a curve C.
(b) 3-space
In other words, as the parameter t varies, we can envision the curve C being traced out by the
moving arrowhead of r(t).
FIGURE 9.1.1 Curves defined by vector We have already seen an example of parametric equations, as well as the vector function of
functions a space curve, in Section 7.5, when we discussed the line in 3-space.

EXAMPLE 1 Circular Helix

Graph the curve traced by the vector function

r(t)  2 cos t i  2 sin t j  t k, t  0.

SOLUTION The parametric equations of the curve are x  2 cos t, y  2 sin t, z  t. By

eliminating the parameter t from the first two equations:

x 2  y2  (2 cos t)2  (2 sin t)2  22

we see that a point on the curve lies on the circular cylinder x2  y2  4. As seen in FIGURE 9.1.2
and the accompanying table, as the value of t increases, the curve winds upward in a spiral
or circular helix.
z
t x y z cylinder
x2 + y2 = 4
0 2
0
0
2
0
( 0, 2,
9π
2 (
( (
p/2 p/2 7π
p 2 0 p 0, –2, (–2, 0, 3π )
2
3p/2 0 2 3p/2
2p 2 0 2p (2, 0, 4 π )
( 0, 2,
5π
2 (
5p/2
3p
0
2
2
0
5p/2
3p
(0, –2,
3π
2 ( (–2, 0, π)

7p/2
4p
0
2
2
0
7p/2
4p
(2, 0, 2 π ) ( 0, 2,
π
2 ( y
9p/2 0 2 9p/2
(2, 0, 0)
x

FIGURE 9.1.2 Circular helix in Example 1

480 | CHAPTER 9 Vector Calculus

 The curve in Example 1 is a special case of the vector function

r(t)  a cos t i  b sin t j  ct k, a  0, b  0, c  0,

which describes an elliptical helix. When a  b, the helix is circular. The pitch of a helix is defined
to be the number 2pc. Problems 9 and 10 in Exercises 9.1 illustrate two other kinds of helixes.

z
EXAMPLE 2 Circle in a Plane
x2 + y2 = 4, z = 3
Graph the curve traced by the vector function

r(t)  2 cos t i  2 sin t j  3 k.

y
SOLUTION The parametric equations of this curve are x  2 cos t, y  2 sin t, z  3. As
x
in Example 1, we see that a point on the curve must also lie on the cylinder x 2  y 2  4.
However, since the z-coordinate of any point has the constant value z  3, the vector function
FIGURE 9.1.3 Curve in Example 2 r(t) traces out a circle 3 units above the xy-plane. See FIGURE 9.1.3.

EXAMPLE 3 Curve of Intersection

z Find the vector function that describes the curve C of intersection of the plane y  2x and the
paraboloid z  9  x 2  y 2.
z = 9 – x2 – y2
SOLUTION We first parameterize the curve C of intersection by letting x  t. It follows
that y  2t and z  9  t 2  (2t)2  9  5t 2. From the parametric equations x  t, y  2t,
C z  9  5t 2, we see that a vector function describing the trace of the paraboloid in the plane
y  2x is given by r(t)  t i  2t j  (9  5t 2) k. See FIGURE 9.1.4.

Limits, Continuity, and Derivatives The fundamental notion of the limit of a vec-
y tor function r(t)   f (t), g(t), h(t) is defined in terms of the limits of the component functions.

x2 + y2 = 9 y = 2x
x Definition 9.1.1 Limit of a Vector Function

FIGURE 9.1.4 Curve in Example 3 If limtSa f (t), limtSa g(t), and limtSa h(t) exist, then
lim r (t)  hlim f (t), lim g(t), lim h(t)i.
tSa tSa tSa tSa

The notation t S a in Definition 9.1.1 can, of course, be replaced by t S a, t S a, t S q, or

t S q.
As an immediate consequence of Definition 9.1.1, we have the following result.

Theorem 9.1.1 Properties of Limits

If limtSa r1(t)  L1 and limtSa r2(t)  L2, then
(i) lim cr1(t)  cL1, c a scalar
tS a
(ii) lim [r1(t)  r2(t)]  L1  L2
tS a
(iii) lim r1(t)  r2(t)  L1  L2.
tS a

Definition 9.1.2 Continuity of a Vector Function

A vector function r is said to be continuous at t  a if
(i) r(a) is defined, (ii) lim r(t) exists, and (iii) lim r(t)  r(a).
tS a tS a

Equivalently, r(t) is continuous at t  a if and only if the component functions f, g, and h are
continuous there.

9.1 Vector Functions | 481

 Definition 9.1.3 Derivative of Vector Function
The derivative of a vector function r is

1
r (t)  lim [r(t  t)  r(t)] (2)
DtS0 Dt

for all t for which the limit exists.

The derivative of r is also written dr/dt. The next theorem will show that on a practical level
the derivative of a vector function is obtained by simply differentiating its component functions.

Theorem 9.1.2 Differentiation of Components

If r(t)   f (t), g(t), h(t), where f, g, and h are differentiable, then

r (t)   f (t), g (t), h (t).

P
tangent PROOF: From (2) we have
Δr
1
r9(t)  lim B k f ( t  Dt), g(t  Dt), h(t  Dt)l 2 k f ( t), g(t), h(t)l R
r(t) DtS0 Dt
z
r(t + Δt) C
f (t  Dt) 2 f (t) g(t  Dt) 2 g(t) h(t  Dt) 2 h(t)
y  lim h , , i.
DtS0 Dt Dt Dt
x
(a)
Taking the limit of each component yields the desired result.
tangent
P
Δr
Smooth Curves When the component functions of a vector function r have continuous
, Δt > 0 first derivatives and r (t) 0 for all t in the open interval (a, b), then r is said to be a smooth
Δt
r(t) function and the curve C traced by r is called a smooth curve.
z
r(t + Δt)
C Geometric Interpretation of r (t) If the vector r (t) is not 0 at a point P, then it may
y be drawn tangent to the curve at P. As seen in FIGURE 9.1.5, the vectors

x Dr 1
(b) r  r(t  t)  r(t) and  [r(t  t)  r(t)]
Dt Dt
FIGURE 9.1.5 Vector r (t) is tangent to
curve C at P are parallel. If we assume lim tS0 r/ t exists, it seems reasonable to conclude that as t S 0,
r(t) and r(t  t) become close and, as a consequence, the limiting position of the vector r/ t
is the tangent line at P. Indeed, the tangent line at P is defined as that line through P parallel to r (t).

r′ (( π
6
y
EXAMPLE 4 Tangent Vectors
Graph the curve C that is traced by a point P whose position is given by r(t)  cos 2t i  sin t j,
r′(0)
0  t  2p. Graph r (0) and r (p/6).
( ( 1 1
,
2 2 SOLUTION By clearing the parameter from the parametric equations x  cos 2t, y  sin t,
(1, 0)
x 0  t  2p, we find that C is the parabola x  1  2y2, 1  x  1. From r (t)  2 sin 2t i  cos t j
we find
x = 1 – 2y2
p !3
r (0)  j and r9 a b  !3 i  j.
6 2

FIGURE 9.1.6 Tangent vectors in In FIGURE 9.1.6 these vectors are drawn tangent to the curve C at (1, 0) and ( 12 , 12 ), respec-
Example 4 tively.

482 | CHAPTER 9 Vector Calculus

 EXAMPLE 5 Tangent Line
Find parametric equations of the tangent line to the graph of the curve C whose parametric
equations are x  t 2, y  t 2  t, z  7t at t  3.
SOLUTION The vector function that gives the position of a point P on the curve is given by
r(t)  t 2 i  (t 2  t) j  7t k. Now,

r (t)  2t i  (2t  1) j  7 k and so r (3)  6 i  5 j  7 k,

which is tangent to C at the point whose position vector is

r(3)  9 i  6 j  21 k;

that is, P(9, 6, 21). Using the components of r (3), we see that parametric equations of the
tangent line are x  9  6t, y  6  5t, z  21  7t.

Higher-Order Derivatives Higher-order derivatives of a vector function are also

obtained by differentiating its components. In the case of the second derivative, we have

r (t)   f (t), g (t), h (t)  f (t) i  g (t) j  h (t) k.

EXAMPLE 6 Derivative of a Vector Function

If r(t)  (t  2t ) i  4t j  et k, then
3 2

r (t)  (3t 2  4t) i  4 j  et k and r (t)  (6t  4) i  et k.

Theorem 9.1.3 Chain Rule

If r is a differentiable vector function and s  u(t) is a differentiable scalar function, then the
derivative of r(s) with respect to t is
dr dr ds
  r (s) u (t).
dt ds dt

EXAMPLE 7 Chain Rule

If r(s)  cos 2s i  sin 2s j  e3s k, where s  t 4, then

dr
 [2 sin 2s i  2 cos 2s j  3e3s k]4t 3
dt
4
 8t 3 sin(2t 4) i  8t 3 cos(2t 4) j  12t 3 e23t k.

Details of the proof of the next theorem are left as exercises.

Theorem 9.1.4 Rules of Differentiation

Let r1 and r2 be differentiable vector functions and u(t) a differentiable scalar function.
d
(i) fr (t)  r2(t)g  r9(t)  r9(t)
dt 1 1 2

d
(ii) fu(t) r1(t)g  u(t) r19(t)  u9(t) r1(t)
dt
d
(iii) fr (t)  r2(t)g  r1(t)  r29(t)  r19(t)  r2(t)
dt 1
d
(iv) fr (t) 3 r2(t)g  r1(t) 3 r29(t)  r19(t) 3 r2(t).
dt 1

9.1 Vector Functions | 483

 Note of caution. Since the cross product of two vectors is not commutative, the order in which r1 and r2 appear
in part (iv) of Theorem 9.1.4 must be strictly observed.

Integrals of Vector Functions If f, g, and h are integrable, then the indefinite and
definite integrals of a vector function r(t)  f (t) i  g(t) j  h(t) k are defined, respectively, by

#r (t) dt  B #f (t) dtR i  B #g(t) dtR j  B #h(t) dtR k

b b b b

#a
r (t) dt  B #a
f (t) dtR i  B #
a
g(t) dtR j  B # h(t) dtR k.
a

The indefinite integral of r is another vector function R  c such that R (t)  r(t).

EXAMPLE 8 Integral of a Vector Function

If r(t)  6t 2 i  4e2t j  8 cos 4t k,

then #r(t) dt  B #6t 2

#
dtR i  B 4e 2t dtR j  B 8 cos 4t dtR k #
 [2t 3  c1] i  [2e2t  c2] j  [2 sin 4t  c3] k

 2t 3 i  2e2t j  2 sin 4t k  c,

where c  c1i  c2 j  c3 k.

Length of a Space Curve If r(t)  f (t) i  g(t) j  h(t) k is a smooth function, then it
can be shown that the length of the smooth curve traced by r is given by
b b
s # "f f 9(t)g
a
2 2
 fg9(t)g  fh9(t)g dt  2
# ir9(t)i dt.
a
(3)

Arc Length as a Parameter A curve in the plane or in space can be parameterized in

terms of the arc length s.

EXAMPLE 9 Example 1 Revisited

Consider the helix of Example 1. Since i r (t) i  !5, it follows from (3) that the length of
the curve from r(0) to an arbitrary point r(t) is
t
s # "5 du  "5t,
0

where we have used u as a dummy variable of integration. Using t  s/ !5, we obtain a vector
equation of the helix as a function of arc length:
s s s
r(s)  2 cos i  2 sin j k. (4)
"5 "5 "5
Parametric equations of the helix are then
s s s
f (s)  2 cos , g(s)  2 sin , h(s)  .
"5 "5 "5
The derivative of a vector function r(t) with respect to the parameter t is a tangent vector to the
curve traced by r. However, if the curve is parameterized in terms of arc length s, then r (s) is a
unit tangent vector. To see this, let a curve be described by r(s), where s is arc length. From (3),
the length of the curve from r(0) to r(s) is s  0s i r (u) i du. Differentiation of this last equation
with respect to s then yields i r (s) i  1.

484 | CHAPTER 9 Vector Calculus

 9.1 Exercises Answers to selected odd-numbered problems begin on page ANS-20.
In Problems 1–10, graph the curve traced by the given vector d
29. [r(t)  (r (t) r (t))]
function. dt
1. r(t)  2 sin t i  4 cos t j  t k; t  0 d
2. r(t)  cos t i  t j  sin t k; t  0 30. [r (t) (r2(t) r3(t))]
dt 1
3. r(t)  t i  2t j  cos t k; t  0
d 1
4. r(t)  4i  2 cos t j  3 sin t k 31. cr1(2t)  r2 a b d
dt t
5. r(t)  et, e2t 
6. r(t)  cosh t i  3 sinh t j d 3 2
32. ft r(t )g
7. r(t)   !2 sin t,!2 sin t, 2 cos t; 0  t  p/2 dt
8. r(t)  t i  t 3 j  t k In Problems 33–36, evaluate the given integral.
9. r(t)  e t cos t i  et sin t j  e t k 2
10. r(t)  t cos t, t sin t, t 2  33. #21
(t i  3t 2 j  4t 3 k) dt
4
In Problems 11–14, find the vector function that describes the
curve C of intersection between the given surfaces. Sketch the 34. # ( "2t  1i 2 "t j  sin p t k) dt
0
curve C. Use the indicated parameter.
11. z  x 2  y2, y  x; x  t 35. # (te i  e j  te k) dt
t 2t t2

12. x 2  y2  z2  1, y  2x; x  t

# 1  t (i  t j  t k) dt
1
13. x 2  y2  9, z  9  x2; x  3 cos t 36. 2
2
14. z  x 2  y2, z  1; x  sin t
sin 2t In Problems 37–40, find a vector function r that satisfies the
15. Given that r(t)  i  (t  2)5j  t ln t k, find
t indicated conditions.
lim 1 r( t).
tS 0
37. r (t)  6i  6t j  3t 2 k; r(0)  i  2j  k
16. Given that limtSa r1(t)  i  2 j  k and limtSa r2(t)  2 i  3
5 j  7 k, find: 38. r (t)  t sin t 2 i  cos 2t j; r(0)  2 i
(a) lim [4r1(t)  3r2(t)] 39. r (t)  12t i  3t 1/2j  2k; r (1)  j, r (1)  2i  k
tS a
40. r (t)  sec2 t i  cos t j  sin t k;
(b) lim r1(t)  r2(t). r (0)  i  j  k, r(0)  j  5k
tS a

In Problems 17–20, find r (t) and r (t) for the given vector function. In Problems 41–44, find the length of the curve traced by the
17. r(t)  ln t i  j, t  0 given vector function on the indicated interval.
18. r(t)  t cos t  sin t, t  cos t  41. r(t)  a cos t i  a sin t j  ct k; 0  t  2p
19. r(t)  te2t, t 3, 4t 2  t 42. r(t)  t i  t cos t j  t sin t k; 0  t  p
20. r(t)  t 2 i  t 3 j  tan1t k 43. r(t)  et cos 2t i  et sin 2t j  et k; 0  t  3p
2
44. r(t)  3t i  !3t 2 j  3 t 3 k; 0  t  1
In Problems 21–24, graph the curve C that is described by r and 45. Express the vector equation of a circle r(t)  a cos t i 
graph r at the indicated value of t. a sin t j as a function of arc length s. Verify that r (s) is a
21. r(t)  2 cos t i  6 sin t j; t  p/6 unit vector.
22. r(t)  t 3 i  t 2 j; t  1 46. If r(s) is the vector function given in (4), verify that r (s) is a
4 unit vector.
23. r(t)  2 i  t j  k; t  1
1  t2 47. Suppose r is a differentiable vector function for which
24. r(t)  3 cos t i  3 sin t j  2t k; t  p/4 i r(t) i  c for all t. Show that the tangent vector r (t) is
perpendicular to the position vector r(t) for all t.
In Problems 25 and 26, find parametric equations of the tangent 48. In Problem 47, describe geometrically the kind of curve C for
line to the given curve at the indicated value of t. which i r(t) i  c.
1 1
25. x  t, y  2 t 2, z  3 t 3; t  2
6t
26. x  t 3  t, y  , z  (2t  1)2; t  1
t11 Miscellaneous Problems
49. Prove Theorem 9.1.4(ii).
In Problems 27–32, find the indicated derivative. Assume that all 50. Prove Theorem 9.1.4(iii).
vector functions are differentiable. 51. Prove Theorem 9.1.4(iv).
d d 52. If v is a constant vector and r is integrable on [a, b], prove
27. [r(t) r (t)] 28. [r(t)  (t r(t))]
dt dt that ab v  r(t) dt  v  ab r(t) dt.

9.1 Vector Functions | 485

 9.2 Motion on a Curve

INTRODUCTION In the preceding section we saw that the first and second derivatives of the
vector function r(t)   f (t), g(t), h(t)  f (t) i  g(t) j  h(t) k can be obtained by differentiat-
ing the component functions f, g, and h. In this section we give a physical interpretation to the
vectors r (t) and r (t).

Velocity and Acceleration Suppose a particle or body moves along a curve C so that
its position at time t is given by the vector function r(t)  f (t) i  g(t) j  h(t) k. If the component
functions f, g, and h have second derivatives, then the vectors

v(t)  r (t)  f (t) i  g (t) j  h (t) k

a(t)  r (t)  f (t) i  g (t) j  h (t) k

are called the velocity and acceleration of the particle, respectively. The scalar function i v(t) i
is the speed of the particle. Since

dr dx 2 dy 2 dz 2
iv(t)i  g g  a b  a b  a b
dt Å dt dt dt
z t
speed is related to arc length s by s (t)  i v(t) i . In other words, arc length is given by s  et01 i v(t) i dt.
v(2)
It also follows from the discussion of Section 9.1 that if P(x1, y1, z1) is the position of the particle
C on C at time t1, then we may draw v(t1) tangent to C at P. Similar remarks hold for curves traced
P(4, 2, 5) by the vector function r(t)  f (t) i  g(t) j.

a(2)
EXAMPLE 1 Velocity and Acceleration Vectors
y The position of a moving particle is given by r(t)  t 2 i  t j  52 t k. Graph the curve defined
by r(t) and the vectors v(2) and a(2).
x = y2
SOLUTION Since x  t 2, y  t, the path of the particle is above the parabola x  y2. When
t  2 the position vector r(2)  4 i  2 j  5 k indicates that the particle is at the point
x
(4, 2, 0) P(4, 2, 5). Now,
5
v(t)  r (t)  2t i  j  k and a(t)  r (t)  2 i
FIGURE 9.2.1 Velocity and acceleration 2
vectors in Example 1 so that v(2)  4 i  j  52 k and a(2)  2 i. These vectors are shown in FIGURE 9.2.1.
If a particle moves with a constant speed c, then its acceleration vector is perpendicular to
the velocity vector v. To see this, note that i v i 2  c2 or v  v  c2. We differentiate both sides
with respect to t and obtain, with the aid of Theorem 9.1.4(iii):
d dv dv dv
(v  v)  v    v  2v   0.
dt dt dt dt
dv
Thus, v0 or a(t)  v(t)  0 for all t.
dt

EXAMPLE 2 Velocity and Acceleration Vectors

Suppose the vector function in Example 2 of Section 9.1 represents the position of a particle
moving in a circular orbit. Graph the velocity and acceleration vector at t  p/4.
SOLUTION Recall that r(t)  2 cos t i  2 sin t j  3 k is the position vector of a particle
moving in a circular orbit of radius 2 in the plane z  3. When t  p/4 the particle is at the
point P( !2, !2, 3). Now,
v(t)  r (t)  2 sin t i  2 cos t j
a(t)  r (t)  2 cos t i  2 sin t j.

486 | CHAPTER 9 Vector Calculus

 Since the speed is i v(t) i  2 for all time t, it follows from the discussion preceding this
z a ((
π
4 example that a(t) is perpendicular to v(t). (Verify this.) As shown in FIGURE 9.2.2, the vectors
v ((
π
4 p p p
v a b  2 sin i  2 cos j  "2 i  "2 j
P(√2 , √2 , 3) 4 4 4

z=3
p p p
y a a b  2 cos i 2 2 sin j  "2 i 2 "2 j
4 4 4
are drawn at the point P. The vector v(p/4) is tangent to the circular path and a(p/4) points
x along a radius toward the center of the circle.
FIGURE 9.2.2 Velocity and acceleration
vectors in Example 2
Centripetal Acceleration For circular motion in the plane, described by r(t)  r0 cos vt i 
r0 sin vt j, r0 and v constants, it is evident that r  v2 r. This means that the acceleration
vector a(t)  r (t) points in the direction opposite to that of the position vector r(t). We then say
v(t1) a(t) is centripetal acceleration. See FIGURE 9.2.3. If v  i v(t) i and a  i a(t) i , we leave it as
an exercise to show that a  v2 /r0.
v(t2)
Curvilinear Motion in the Plane Many important applications of vector functions
r(t2) a(t1)
occur in regard to curvilinear motion in a plane. For example, planetary and projectile motion
take place in a plane.
In analyzing the motion of short-range ballistic projectiles,* we begin with the acceleration
of gravity written in vector form: a(t)  g j.
If, as shown in FIGURE 9.2.4, a projectile is launched with an initial velocity v0  v0 cos u i 
FIGURE 9.2.3 Circular motion v0 sin u j from an initial height s0  s0 j, then

y #
v(t)  (g j) dt  gt j  c1,
v0 where v(0)  v0 implies that c1  v0. Therefore,
(v0 sin θ ) j
θ v(t)  (v0 cos u) i  (gt  v0 sin u) j.
(v0 cos θ )i
s0 j Integrating again and using r(0)  s0 yields
x
1
r(t)  (v0 cos u)t i  c gt 2  (v0 sin u) t  s0 d j. (1)
2
FIGURE 9.2.4 Trajectory of a projectile
Hence, parametric equations for the trajectory of the projectile are
y 1
x(t)  (v0 cos u)t, y(t)   gt 2  (v0 sin u)t  s0. (2)
2
We are naturally interested in finding the maximum height H and the range R attained by a
H
projectile. As shown in FIGURE 9.2.5, these quantities are the maximum values of y(t) and x(t),
respectively.
x
EXAMPLE 3 Trajectory of a Projectile
(a) Maximum height H: A projectile is launched from ground level with an initial speed v0  768 ft/s at an angle of
Find t1 for which y′(t1) = 0; elevation u  30. Find (a) the vector function and parametric equations of the projectile’s
H = ymax = y(t1)
y
trajectory, (b) the maximum altitude attained, (c) the range of the projectile, and (d) the
impact speed.
SOLUTION (a) The initial height and velocity are, respectively, s0  0 and

r9(0)  v0  (768 cos 308)i  (768 sin 308)j  384"3i  384j. (3)
R
x
Integrating a(t)  32j and using (3) give
(b) Range R:
Find t1 > 0 for which y (t1) = 0; v(t)  (384"3)i  (32t  384)j. (4)
R = xmax = x(t1)

FIGURE 9.2.5 Maximum height and range *A projectile is shot or hurled rather than self-propelled. In the analysis of long-range ballistic motion,
of a projectile the curvature of the Earth must be taken into consideration.

9.2 Motion on a Curve | 487

 Integrating (4) and using r(0)  s0  0 then give

r(t)  (384"3)t i  (16t 2  384t)j.

The components of this vector function,

x(t)  (384"3)t, y(t)  16t 2  384t, (5)

are the parametric equations of the projectile’s trajectory.

(b) From (5) we see that y (t)  0 when 32t  384  0 or t  12 s. The maximum
height attained by the projectile is

H  y(12)  16(12)2  384(12)  2304 ft.

(c) From (5) we see that y(t)  0 when 16t 2  384t  16t(t 2 24)  0. The time
that projectile hits the ground is t  24 s and the corresponding range is

R  x(24)  384"3(24) < 15,963 ft.

(d) Finally, from (4) we see that v(24)  (384"3)i  (384)j and so the impact speed is

7 v(24) 7  "(384!3)2  (384)2  768 ft/s.

In Example 3, note that the impact speed is the same as the launch speed v0  768 ft/s. Verify
that this is still the case if we change the angle of elevation to, say, u  50. See Problem 16 in
Exercises 9.2.

REMARKS
We have seen that the rate of change of arc length ds/dt is the same as the speed i v(t)i  i r (t)i .
However, as we shall see in the next section, it does not follow that the scalar acceleration
d 2s/dt 2 is the same as i a(t)i  i r (t)i . See Problem 24 in Exercises 9.2.

9.2 Exercises Answers to selected odd-numbered problems begin on page ANS-21.

In Problems 1–8, r(t) is the position vector of a moving particle. pass through the xy-plane? What are its velocity and accel-
Graph the curve and the velocity and acceleration vectors at the eration at these points?
indicated time. Find the speed at that time. 10. Suppose a particle moves in space so that a(t)  0 for all time t.
1
1. r(t)  t 2 i  4 t 4 j; t  1
Describe its path.
11. A shell is fired from ground level with an initial speed of
1
2. r(t)  t 2 i  j; t  1 480 ft/s at an angle of elevation of 30. Find:
t2
3. r(t)  cosh 2t i  sinh 2t j; t  0 (a) a vector function and parametric equations of the shell’s
4. r(t)  2 cos t i  (1  sin t) j; t  p/3 trajectory,
5. r(t)  2 i  (t  1)2 j  t k; t  2 (b) the maximum altitude attained,
6. r(t)  t i  t j  t 3 k; t  2 (c) the range of the shell, and
7. r(t)  t i  t 2 j  t 3 k; t  1 (d) the speed at impact.
8. r(t)  t i  t 3 j  t k; t  1 12. Rework Problem 11 if the shell is fired with the same initial
9. Suppose r(t)  t 2 i  (t 3  2t) j  (t 2  5t) k is the position speed and the same angle of elevation but from a cliff 1600 ft
vector of a moving particle. At what points does the particle high.

488 | CHAPTER 9 Vector Calculus

13. A used car is pushed off an 81-ft-high sheer seaside cliff with 19. If a projectile is launched from level ground, the initial con-
a speed of 4 ft/s. Find the speed at which the car hits the water. ditions are r(0)  0, r9(0)  v0  (v0 cos u)i  (v0 sin u)j,
14. A small projectile is launched from ground level with an initial and if linear air resistance is taken into consideration, the
speed of 98 m/s. Find the possible angles of elevation so that vector function analogue of (1) is
its range is 490 m.
15. A football quarterback throws a 100-yd “bomb” at an angle mv0 cos u
of 45 from the horizontal. What is the initial speed of the r(t)  (1 2 e bt>m) i
b
football at the point of release? mv0 sin u m2g mg
16. If a projectile is launched from level ground, then the initial condi-  ca  2 b(1 2 e bt>m) 2 td j,
b b b
tions are r(0)  0, r9(0)  v0  (v0 cos u)i  (v0 sin u)j and
the vector function (1) becomes This function is equivalent to the solution x(t), y(t) of the
system of linear differential equations (11) in Problem 22 of
r(t)  (v0 cos u)t i  f12gt 2  (v0 sin u)tg j. (6) Exercises 4.6 subject to the same initial conditions. Here b . 0
is a constant of proportionality related to the air resistance or
This function is equivalent to the solution x(t), y(t) of the drag. If m  14 slug, g  32 ft/s 2, b  0.02, v0  300 ft/s,
system of linear differential equations (8) in Problem 21 of and u  38, use a calculator or CAS to find the impact speed
Exercises 4.6 subject to the same initial conditions. Underlying of the projectile. See part (b) of Problem 22 in Exercises 4.6.
that earlier discussion, as well as in the derivation of (1), is the 20. Suppose the angle of elevation of the projectile in Problem 19
assumption that the projectile is not subject to air resistance. is changed to u  52. Do you think that the impact speed of
Use (6) to show that the impact speed of a projectile is the the projectile is the same, greater than, or less than the value
same as the initial speed 7 r9(0) 7  7 v0 7  v0 for any angle found in that problem? Prove your assertion.
of elevation u, 0 , u , 90. 21. A projectile is fired from a cannon directly at a target that
17. When air resistance is ignored, we saw in Problem 21 is dropped from rest simultaneously as the cannon is fired.
of Exercises 4.6 that a projectile launched at an angle u, Show that the projectile will strike the target in midair. See
0 , u , 90 from level ground (r(0)  0), then its horizontal FIGURE 9.2.6. [Hint: Assume that the origin is at the muzzle of
range and maximum height are, respectively, the cannon and that the angle of elevation is u. If rp and rt are
position vectors of the projectile and target, respectively, is
v 20 v 20 there a time at which rp  rt?]
R sin 2u and H sin2 u.
g 2g

From the first formula, it follows that when the projectile

is launched at distinct complementary angles the horizontal
range is the same and that the maximum range is attained
when the angle of elevation is u  45. But if the projectile is
launched from an initial height s0 . 0 the foregoing formulas
and statements are not valid.
(a) With v0  480 ft/s, s0  1600 ft rework Problem 12,
but this time use the complementary angle of elevation
u  60. Compare the maximum height, range, and im- FIGURE 9.2.6 Cannon in Problem 21
pact speed of the projectile with the answers to parts (b),
(c) and (d) of Problem 12. 22. In army field maneuvers sturdy equipment and supply packs
(b) Use a graphing utility or CAS to plot the trajectory of the are simply dropped from planes that fly horizontally at a
projectile defined by the parametric equations x(t) and slow speed and a low altitude. A supply plane flies hori-
y(t) in part (a) of Problem 12. Repeat for the parametric zontally over a target at an altitude of 1024 ft at a constant
equations in part (a) of this problem. Superimpose both speed of 180 mi/h. Use (1) to determine the horizontal dis-
of these curves on the same coordinate system. tance a supply pack travels relative to the point from which
18. As mentioned in Problem 17, if a projectile is launched from it was dropped. At what line-of-sight angle a should the
an initial height s0 . 0 the maximum range of the projectile supply pack be released in order to hit the target indicated
is not attained using u  45 as the angle of elevation. in FIGURE 9.2.7?
(a) A projectile is launched from an initial height of v0  480 ft/s,
s0  1600 ft, at an angle of elevation of u  45. Use a
calculator or CAS to find the time the projectile hits the α supply
ground and the corresponding range. pack
1024 ft
(b) If the angle of elevation in part (a) is changed to u  39.76,
show that the range is greater than that in part (a).
(c) Use a graphing utility or CAS to plot the trajectories of target
the projectiles in parts (a) and (b). Superimpose both of
these curves on the same coordinate system. FIGURE 9.2.7 Supply plane in Problem 22

9.2 Motion on a Curve | 489

23. Suppose that r(t) ⫽ r0 cos vt i ⫹ r0 sin vt j is the position 29. Suppose the Sun is located at the origin. The gravitational force
vector of an object that is moving in a circle of radius r0 in F exerted on a planet of mass m by the Sun of mass M is
the xy-plane. If i v(t) i ⫽ v, show that the magnitude of the Mm
centripetal acceleration is a ⫽ i a(t) i ⫽ v2 /r0. F 5 2k u.
r2
24. The motion of a particle in space is described by
F is a central force—that is, a force directed along the posi-
r(t) ⫽ b cos t i ⫹ b sin t j ⫹ ct k, t ⱖ 0. tion vector r of the planet. Here k is the gravitational constant,
r ⫽ i r i , u ⫽ r/r is a unit vector in the direction of r, and the
(a) Compute i v(t) i . minus sign indicates that F is an attractive force—that is, a
(b) Compute s ⫽ 兰0t i v(t) i dt and verify that ds/dt is the same force directed toward the Sun. See FIGURE 9.2.9.
as the result of part (a). (a) Use Problem 28 to show that the torque acting on the
(c) Verify that d 2s/dt 2 ⫽ i a(t) i . planet due to this central force is 0.
25. The effective weight we of mass m at the equator of the Earth (b) Explain why the angular momentum L of a planet is constant.
is defined by we ⫽ mg ⫺ ma, where a is the magnitude of the
planet
centripetal acceleration given in Problem 23. Determine the
m
effective weight of a 192-lb person if the radius of the Earth sun
is 4000 mi, g ⫽ 32 ft/s2, and v ⫽ 1530 ft/s. r F
26. Consider a bicyclist riding on a flat circular track of radius
M
r0. If m is the combined mass of the rider and bicycle, fill in
the blanks in FIGURE 9.2.8. [Hint: Use Problem 23 and force ⫽
mass ⫻ acceleration. Assume that the directions are upward FIGURE 9.2.9 Force F in Problem 29
and to the left.] The resultant vector U gives the direction the
bicyclist must be tipped to avoid falling. Find the angle f from
the vertical at which the bicyclist must be tipped if her speed Discussion Problems
is 44 ft/s and the radius of the track is 60 ft. 30. In this problem the student will use the properties in Sections
resultant
7.4 and 9.1 to prove Kepler’s first law of planetary motion:
〈0, —〉 The orbit of a planet is an ellipse with the Sun at one focus. We
U = 〈—, —〉
force exerted assume that the Sun has mass M and is located at the origin, r
by track = is the position vector of a body of mass m moving under the
opposite of
the combined gravitational attraction of the Sun, and u ⫽ r/r is a unit vector
φ
weight of bike in the direction of r.
and person
(a) Use Problem 29 and Newton’s second law of motion
F ⫽ m a to show that

〈—, 0〉 d 2r u
5 2kM 2 .
dt 2 r
centripetal force
(b) Use part (a) to show that r ⫻ r⬙ ⫽ 0.
d
(c) Use part (b) to show that (r ⫻ v) ⫽ 0.
dt
FIGURE 9.2.8 Bicyclist in Problem 26 (d) It follows from part (c) that r ⫻ v ⫽ c, where c is a constant
vector. Show that c ⫽ r 2 (u ⫻ u⬘).
27. The velocity of a particle moving in a fluid is described by d
means of a velocity field v ⫽ v1 i ⫹ v2 j ⫹ v3 k, where the com- (e) Show that (u ⭈ u) ⫽ 0 and consequently u ⭈ u⬘ ⫽ 0.
dt
ponents v1, v2, and v3 are functions of x, y, z, and time t. If the (f ) Use parts (a), (e), and (d) to show that
velocity of the particle is v(t) ⫽ 6t 2 x i ⫺ 4ty2 j ⫹ 2t(z ⫹ 1) k,
find r(t). [Hint: Use separation of variables.] d du
(v ⫻ c) ⫽ kM .
28. Suppose m is the mass of a moving particle. Newton’s second dt dt
law of motion can be written in vector form as
(g) By integrating the result in part (f) with respect to t, we
d dp get v ⫻ c ⫽ kMu ⫹ d, where d is another constant vector.
F ⫽ m a ⫽ (m v) ⫽ , Dot both sides of this last expression by the vector r ⫽
dt dt
r u and use Problem 61 in Exercises 7.4 to show that
where p ⫽ m v is called linear momentum. The angular
momentum of the particle with respect to the origin is defined c 2>kM
r⫽ ,
to be L ⫽ r ⫻ p, where r is its position vector. If the torque 1 ⫹ (d>kM) cos u
of the particle about the origin is t ⫽ r ⫻ F ⫽ r ⫻ d p/dt,
show that t is the time rate of change of angular momentum. where c ⫽ i c i , d ⫽ i d i , and u is the angle between d and r.

490 | CHAPTER 9 Vector Calculus

 (h) Explain why the result in part (g) proves Kepler’s first law. Note that when s0  0 this formula reduces to the range R
(i) At perihelion, the point in the orbit where the body is given in Problem 17.
closest to the Sun, the vectors r and v are perpendicular 32. Consider the formula in Problem 31 as a function of the single
and have magnitudes r0 and v0, respectively. Use this variable u. Show that the range of a projectile launched from
information and parts (d) and (g) to show that c  r0v0 an initial height s0 is a maximum for the angle of inclination
and d  r0v02  kM.
31. Suppose a projectile is launched from an initial height s0 at 2s0g  v 20
an angle of elevation u. If air resistance is ignored, show that u  cos1 .
Å 2s0g  2v 20
the horizontal range is given by
Note that when s0  0 this formula reduces to the value given
v0 cos u in Problem 17, that is, u  p/4 or 45. Use this formula to
R (v0 sin u  "v 20 sin2 u  2s0g ).
g verify the angle of inclination used in part (b) of Problem 18.

9.3 Curvature and Components of Acceleration

INTRODUCTION Let C be a smooth curve in either 2- or 3-space traced out by a vector

function r(t). In this section we are going to consider in greater detail the acceleration vector
a(t)  r (t) introduced in the last section. But before doing this, we need to examine a scalar
quantity called the curvature of a curve.

A Definition We know that r (t) is a tangent vector to the curve C, and consequently

r9(t)
T(t)  (1)
ir9(t)i
T
T is a unit tangent. But recall from the end of Section 9.1 that if C is parameterized by arc length s,
T
then a unit tangent to the curve is also given by dr/ds. The quantity i r (t) i in (1) is related to arc
T
C length s by ds/dt  i r (t) i . Since the curve C is smooth, we know from pages 482 and 484 that
T
ds/dt  0. Hence by the Chain Rule,
P3 T
T
P2 dr d r ds dr d r>dt r9(t)
 and so    T(t). (2)
dt ds dt ds ds>dt ir9(t)i
T

Now suppose C is as shown in FIGURE 9.3.1. As s increases, T moves along C, changing direction
T
but not length (it is always of unit length). Along the portion of the curve between P1 and P2 the
vector T varies little in direction; along the curve between P2 and P3, where C obviously bends
P1 more sharply, the change in the direction of the tangent T is more pronounced. We use the rate
at which the unit vector T changes direction with respect to arc length as an indicator of the
FIGURE 9.3.1 Unit tangents curvature of a smooth curve C.

Definition 9.3.1 Curvature

Let r(t) be a vector function defining a smooth curve C. If s is the arc length parameter and
T  d r/ds is the unit tangent vector, then the curvature of C at a point is
dT
k5 g g. (3)
ds

The symbol k in (3) is the Greek letter kappa. Now since curves are generally not parameterized
by arc length, it is convenient to express (3) in terms of a general parameter t. Using the Chain
Rule again, we can write
dT d T ds dT d T>dt
5 and consequently 5 .
dt ds dt ds ds>dt

9.3 Curvature and Components of Acceleration | 491

 In other words, curvature is given by

iT9(t)i
k(t)  . (4)
ir9(t)i

EXAMPLE 1 Curvature of a Circle

Find the curvature of a circle of radius a.
SOLUTION A circle can be described by the vector function r(t)  a cos t i  a sin t j. Now
from r (t)  a sin t i  a cos t j and i r (t) i  a, we get
r9(t)
T(t)   sin t i  cos t j and T (t)  cos t i  sin t j.
ir9(t)i
large curvature κ
Hence, from (4) the curvature is
iT9(t)i " cos2 t  sin2 t 1
k(t)    . (5)
ir9(t)i a a
small curvature κ
The result in (5) shows that the curvature at a point on a circle is the reciprocal of the radius
FIGURE 9.3.2 Curvature of a circle in of the circle and indicates a fact that is in keeping with our intuition: A circle with a small
Example 1 radius curves more than one with a large radius. See FIGURE 9.3.2.

Tangential and Normal Components of Acceleration Suppose a particle moves

in 2- or 3-space on a smooth curve C described by the vector function r(t). Then the velocity of the
particle on C is v(t)  r (t), whereas its speed is ds/dt  v  i v(t) i . Thus, (1) implies v(t)  v T.
Differentiating this last expression with respect to t gives acceleration:
dT dv
a(t)  v  T. (6)
dt dt
z Furthermore, with the help of Theorem 9.1.4(iii), it follows from the differentiation of
aTT
T  T  1 that T  d T/dt  0. Hence, at a point P on C the vectors T and d T/dt are orthogonal.
a
If i d T/dt i 0, the vector
T
d T>dt
P N(t)  (7)
id T>dti
N
is a unit normal to the curve C at P with direction given by d T/dt. The vector N is also called
aNN idT>dti
the principal normal. But since curvature is k  , it follows from (7) that d T/dt 
y v
kv N. Thus, (6) becomes
dv
x a(t)  kv2 N  T. (8)
C dt
By writing (8) as a(t)  aN N  aT T, (9)
FIGURE 9.3.3 Components of
acceleration we see that the acceleration vector a of the moving particle is the sum of two orthogonal vec-
tors aN N and aT T. See FIGURE 9.3.3. The scalar functions aT  dv/dt and aN  kv2 are called the
z
B=T×N tangential and normal components of the acceleration, respectively. Note that the tangential
component of the acceleration results from a change in the magnitude of the velocity v, whereas
T the normal component of the acceleration results from a change in the direction of v.
P
N
The Binormal A third unit vector defined by
osculating plane
y
B(t)  T(t) N(t)
is called the binormal. The three unit vectors T, N, and B form a right-handed set of mutually
x
C orthogonal vectors called the moving trihedral. The plane of T and N is called the osculating
plane,* the plane of N and B is said to be the normal plane, and the plane of T and B is the
FIGURE 9.3.4 Osculating plane rectifying plane. See FIGURE 9.3.4.

*Literally, this means the “kissing” plane.

492 | CHAPTER 9 Vector Calculus

 The three mutually orthogonal unit vectors T, N, and B can be thought of as a movable right-
handed coordinate system since
B(t)  T(t) N(t), N(t)  B(t) T(t), T(t)  N(t) B(t).
This movable coordinate system is referred to as the TNB-frame.

EXAMPLE 2 Tangent, Normal, and Binormal Vectors

The position of a moving particle is given by r(t)  2 cos t i  2 sin t j  3t k. Find the vectors
T, N, and B. Find the curvature.

SOLUTION Since r (t)  2 sin t i  2 cos t j  3 k, i r (t)i  "13, and so from (1) we
see that a unit tangent is

2 2 3
T(t)   sin t i  cos t j  k.
"13 "13 "13
Next, we have

dT 2 2 dT 2
 cos t i 2 sin t j and g g  .
dt "13 "13 dt "13
Hence, (7) gives the principal normal

N(t)  cos t i  sin t j.

Now, the binormal is i j k

2 2 3
B(t)  T(t) 3 N(t)  4  sin t cos t 4
"13 "13 "13
 cos t sin t 0

3 3 2
 sin t i 2 cos t j  k.
"13 "13 "13
Finally, using i d T/dti  2/ "13 and i r (t)i  "13, we obtain from (4) that the curvature
at any point is the constant
2> "13 2
k  .
"13 13

The fact that the curvature in Example 2 is constant is not surprising, since the curve defined
by r(t) is a circular helix.

EXAMPLE 3 Osculating, Normal, Rectifying Planes

At the point corresponding to t  p>2 on the circular helix in Example 2, find an equation of
(a) the osculating plane, (b) the normal plane, and (c) the rectifying plane.
SOLUTION From r(p>2)  k0, 2, 3p>2l the point P in question is (0, 2, 3p>2).
(a) A normal vector to the osculating plane at P is

3 2
B(p>2)  T(p>2) 3 N(p>2)  i k.
"13 "13
To find an equation of a plane we do not require a unit normal, so in lieu of B(p>2) it
is a bit simpler to use k3, 0, 2l. From (11) of Section 7.5 an equation of the osculating
plane is

3p
3(x 2 0)  0(y 2 2)  2az 2 b 0 or 3x  2z  3p.
2

9.3 Curvature and Components of Acceleration | 493

 2y 1
(b) At the point P, the vector T(p>2)  !13 k2, 0, 3l or k2, 0, 3l is normal to the plane
0
–2 containing N(p>2) and B(p>2). Hence an equation of the normal plane is

20 3p
2(x 2 0)  0(y 2 2)  3az 2 b 0 or 4x  6z  9p.
2
10
z (c) Finally, at the point P, the vector N(p>2)  k0, 1, 0l is normal to the plane contain-
0 ing T(p>2) and B(p>2). An equation of the rectifying plane is
–10 3p
0(x 2 0)  (1)(y 2 2)  0az 2 b 0 or y  2.
2
2.5 5
–2.5 0
–5 x With the help of Mathematica, portions of the helix and the osculating plane in Example 3 are
shown in FIGURE 9.3.5. The point (0, 2, 3p>2) is indicated in the figure by the red dot.
FIGURE 9.3.5 Helix and osculating plane
in Example 3 Formulas for aT , aN , and Curvature By dotting, and in turn crossing, the vector
v  v T with (9), it is possible to obtain explicit formulas involving r, r , and r for the tangential
and normal components of the acceleration and the curvature. Observe that

v  a  aN (v T  N)  aT (v T  T)  aTv

0 1

yields the tangential component of acceleration

dv va r9(t)  r0(t)
aT    . (10)
dt ivi ir9(t)i

On the other hand,

v a  aN (v T N)  aT (v T T)  aNv B.

B 0

Since i B i  1, it follows that the normal component of acceleration is

iv 3 ai ir9(t) 3 r0(t)i
aN  kv 2   . (11)
ivi ir9(t)i
Solving (11) for the curvature gives

iv 3 ai ir9(t) 3 r0(t)i
k(t)   . (12)
ivi3 ir9(t)i3

EXAMPLE 4 Curvature of Twisted Cubic

The curve traced by r(t)  t i  12 t 2 j  13 t 3 k is said to be a “twisted cubic.” If r(t) is the position
vector of a moving particle, find the tangential and normal components of the acceleration at
any t. Find the curvature.
SOLUTION v(t)  r (t)  i  t j  t 2 k, a(t)  r (t)  j  2t k.

Since v  a  t  2t 3 and i vi  "1  t 2  t 4 , it follows from (10) that

dv t  2t 3
aT   .
dt "1  t 2  t 4

i j k
Now, v 3 a  31 t t 2 3  t 2 i 2 2t j  k
0 1 2t

494 | CHAPTER 9 Vector Calculus

 and i v ai  "t 4  4t 2  1. Thus, (11) gives

"t 4  4t 2  1 t 4  4t 2  1
aN  kv 2   .
"1  t 2  t 4 Å t4  t2  1

From (12) we find that the curvature of the twisted cubic is given by
(t 4  4t 2  1) 1>2
k(t)  .
(t 4  t 2  1) 3>2
tangent
Radius of Curvature The reciprocal of the curvature, r  1/k, is called the radius of
C curvature. The radius of curvature at a point P on a curve C is the radius of a circle that “fits”
P the curve there better than any other circle. The circle at P is called the circle of curvature and
its center is the center of curvature. The circle of curvature has the same tangent line at P as
ρ the curve C, and its center lies on the concave side of C. For example, a car moving on a curved
track, as shown in FIGURE 9.3.6, can, at any instant, be thought to be moving on a circle of radius r.
Hence, the normal component of its acceleration aN  kv2 must be the same as the magnitude
of its centripetal acceleration a  v2 /r. Therefore, k  1/r and r  1/k. Knowing the radius of
curvature, we can determine the speed v at which a car can negotiate a banked curve without
FIGURE 9.3.6 Radius of curvature skidding. (This is essentially the idea in Problem 26 in Exercises 9.2.)

REMARKS
By writing (6) as
ds d T d 2s
a(t)   2 T,
dt dt dt
we note that the so-called scalar acceleration d 2s/dt 2, referred to in the last remark, is now
seen to be the tangential component of the acceleration aT.

9.3 Exercises Answers to selected odd-numbered problems begin on page ANS-21.

In Problems 1 and 2, for the given position function, find the In Problems 7–16, r(t) is the position vector of a moving
unit tangent. particle. Find the tangential and normal components of the
acceleration at any t.
1. r(t)  (t cos t  sin t) i  (t sin t  cos t) j  t 2 k, t  0
2. r(t)  et cos t i  et sin t j  "2et k
7. r(t)  i  t j  t 2 k
3. Use the procedure outlined in Example 2 to find T, N, B, and
8. r(t)  3 cos t i  2 sin t j  t k
k for motion on a general circular helix that is described by 9. r(t)  t 2 i  (t 2  1) j  2t 2 k
r(t)  a cos t i  a sin t j  ct k. 10. r(t)  t 2 i  t 3 j  t 4 k
4. Use the procedure outlined in Example 2 to show on the twisted
11. r(t)  2t i  t 2 j
cubic of Example 4 that at t  1: 12. r(t)  tan1 t i  12 ln(1  t 2) j
13. r(t)  5 cos t i  5 sin t j
1 1 14. r(t)  cosh t i  sinh t j
T (i  j  k), N   (i 2 k),
"3 "2 15. r(t)  et (i  j  k)
16. r(t)  t i  (2t  1) j  (4t  2) k
1 "2 17. Find the curvature of an elliptical helix that is described by
B (i  2 j 2 k), k  .
"6 3 r(t)  a cos t i  b sin t j  ct k, a  0, b  0, c  0.
18. (a) Find the curvature of an elliptical orbit that is described
In Problems 5 and 6, find an equation of the osculating plane by r(t)  a cos t i  b sin t j  c k, a  0, b  0, c  0.
to the given space curve at the point that corresponds to (b) Show that when a  b, the curvature of a circular orbit
the indicated value of t. is the constant k  1/a.
5. The circular helix of Example 2; t  p/4 19. Show that the curvature of a straight line is the constant k  0.
6. The twisted cubic of Example 4; t  1 [Hint: Use (2) in Section 7.5.]

9.3 Curvature and Components of Acceleration | 495

20. Find the curvature at t  p of the cycloid that is described by In Problems 23 and 24, use the result of Problem 22 to find the
curvature and radius of curvature of the curve at the indicated
r(t)  a(t  sin t) i  a(1  cos t) j, a  0.
points. Decide at which point the curve is “sharper.”
21. Let C be a plane curve traced by r(t)  f (t) i  g(t) j, where f
23. y  x 2; (0, 0), (1, 1)
and g have second derivatives. Show that the curvature at a
24. y  x 3; (1, 1), ( 12 , 18 )
point is given by
Z f9(t)g0(t) 2 g9(t) f 0(t)Z
k . Discussion Problems
(f f9(t)g 2  fg9(t)g 2)3>2
25. Discuss the curvature near a point of inflection of y  F(x).
22. Show that if y  F(x), the formula for k in Problem 21 reduces to 26. Show that i a(t) i 2  a2N  a2T .
ZF0(x)Z
k .
f1  (F9(x))2g 3>2

9.4 Partial Derivatives

z
(x, y, z), where z = f (x, y)
INTRODUCTION In this section we consider functions of two or more variables and how
to find the instantaneous rate of change—that is, the derivative—of such functions with respect
to each variable.

f (x, y) Functions of Two Variables Recall from calculus that a function of two variables
y
is a rule of correspondence that assigns to each ordered pair of real numbers (x, y) of a subset
of the xy-plane one and only one number z in the set R of real numbers. The set of ordered pairs
(x, y)
(x, y) is called the domain of the function and the set of corresponding values of z is called the
x domain of z = f (x, y) range. A function of two variables is usually written z  f (x, y). The variables x and y are called
the independent variables of the function, and z is called the dependent variable. The graph
FIGURE 9.4.1 Function of two variables of a function z  f (x, y) is a surface in 3-space. See FIGURE 9.4.1.

Level Curves For a function z  f (x, y), the curves defined by f (x, y)  c, for suitable
c, are called the level curves of f. The word level arises from the fact that we can interpret the
z equation f (x, y)  c as the projection onto the xy-plane of the curve of intersection, or trace, of
z  f (x, y) and the (horizontal or level) plane z  c. See FIGURE 9.4.2.
z = y2 – x2
c=1 z

y plane
z=c y

x surface
increasing
z = f (x, y) values of f
(a) Surface
y x

y c=0
c=1
x
f (x, y) = c
(a) Surface (b) Level curves
x

c = –1 FIGURE 9.4.2 Surface and level curves

EXAMPLE 1 Level Curves

(b) Level curves
The level curves of the function f (x, y)  y2  x 2 are defined by y2  x 2  c. As shown in
FIGURE 9.4.3 Surface and level curves in FIGURE 9.4.3, when c  0 or c  0, a member of this family of curves is a hyperbola. For c  0,
Example 1 we obtain the lines y  x and y  x.

496 | CHAPTER 9 Vector Calculus

 Functions of Three or More Variables Functions of three or more variables are
defined analogously to functions of two variables. For example, a function of three variables
is a rule of correspondence that assigns to each ordered triple of real numbers (x, y, z) of a subset
z
c=1
of 3-space one and only one number w in the set R of real numbers. We write w  F(x, y, z).

Level Surfaces Although we cannot draw a graph of a function of three variables

w  F(x, y, z), we can draw the surfaces defined by F(x, y, z)  c for suitable values of c. These
c=2 surfaces are called level surfaces. This is an unfortunate, though standard, choice of words, since
level surfaces are usually not level.
y
EXAMPLE 2 Level Surfaces
Describe the level surfaces of the function F(x, y, z)  (x 2  y2)/z.
c = –2
x SOLUTION For c 0 the level surfaces are given by
x2 1 y2
c = –1 c or x 2  y 2  cz.
z
FIGURE 9.4.4 Level surfaces in Example 2 A few members of this family of paraboloids are shown in FIGURE 9.4.4.

Partial Derivatives The derivative of a function of one variable y  f (x) is given by

dy f (x  Dx) 2 f (x)
 lim .
dx DxS0 Dx
In exactly the same manner, we can define a derivative of a function of two variables with respect
to each variable. If z  f (x, y), then the partial derivative with respect to x is
0z f (x  Dx, y) 2 f (x, y)
 lim (1)
0x DxS0 Dx

and the partial derivative with respect to y is

0z f (x, y  Dy) 2 f (x, y)
 lim , (2)
0y DyS0 Dy

provided each limit exists.

In (1) the variable y does not change in the limiting process; that is, y is held fixed. Similarly,
in (2) the variable x is held fixed. The two partial derivatives (1) and (2) then represent the rates
of change of f with respect to x and y, respectively. On a practical level:
To compute z / x, use the laws of ordinary differentiation while treating y as a constant.
To compute z / y, use the laws of ordinary differentiation while treating x as a constant.

EXAMPLE 3 Partial Derivatives

If z  4x 3y2  4x 2  y6  1, find z / x and z / y.
SOLUTION We hold y fixed and treat constants in the usual manner. Thus,
0z
 12x 2y2  8x.
0x
By treating x as a constant, we obtain
0z
 8x 3y  6y5.
0y

Alternative Symbols The partial derivatives z /x and z /y are often represented by
alternative symbols. If z  f (x, y), then
0z 0f 0z 0f
5 5 zx 5 fx and 5 5 zy 5 fy .
0x 0x 0y 0y

9.4 Partial Derivatives | 497

 Higher-Order and Mixed Derivatives For a function of two variables z ⫽ f (x, y),
the partial derivatives ⭸z /⭸x and ⭸z /⭸y are themselves functions of x and y. Consequently, we
can compute second and higher partial derivatives. Indeed, we can find the partial derivative
of ⭸z/⭸x with respect to y, and the partial derivative of ⭸z/⭸y with respect to x. The latter types of
partial derivatives are called mixed partial derivatives. In summary, for z ⫽ f (x, y):
Second-order partial derivatives:
0 2z 0 0z 0 2z 0 0z
⫽ a b and ⫽ a b.
0x 2 0x 0x 0y 2 0y 0y
Third-order partial derivatives:
0 3z 0 0 2z 0 3z 0 0 2z
3
⫽ a b and ⫽ a b.
0x 0x 0x 2 0y 3
0y 0y 2
Mixed second-order partial derivatives:
0 2z 0 0z 0 2z 0 0z
⫽ a b and ⫽ a b.
0x0y 0x 0y 0y0x 0y 0x

Alternative Symbols The second- and third-order partial derivatives are denoted by fxx ,
fyy , fxxx, and so on. The subscript notation for mixed second partial derivatives is fxy or fyx. Note that
0 0z 0 2z 0 2z
fxy ⫽ ( fx)y ⫽ a b ⫽ and fyx ⫽ .
0y 0x 0y0x 0x0y
Although we shall not prove it, if a function f has continuous second partial derivatives, then the
order in which a mixed second partial derivative is done is irrelevant; that is,

fxy ⫽ fyx . (3)

Functions of Three or More Variables The rates of change of a function of three vari-
ables w ⫽ F(x, y, z) in the x, y, and z directions are ⭸w/⭸x, ⭸w/⭸y, and ⭸w/⭸z, respectively. To compute,
say, ⭸w/⭸x, we differentiate with respect to x in the usual manner while holding both y and z constant.
In this manner we extend the process of partial differentiation to functions of any number of variables.

EXAMPLE 4 Partial Derivatives

⫺3pt
If F(x, y, t) ⫽ e cos 4x sin 6y, then the partial derivatives with respect to x, y, and t are, in turn,
Fx(x, y, t) ⫽ ⫺4e⫺3pt sin 4x sin 6y,
Fy(x, y, t) ⫽ 6e⫺3pt cos 4x cos 6y,
Ft(x, y, t) ⫽ ⫺3pe⫺3pt cos 4x sin 6y.

Chain Rule The Chain Rule for functions of one variable states that if y ⫽ f (u) is a dif-
ferentiable function of u, and u ⫽ g(x) is a differentiable function of x, then the derivative of the
composite function is
dy dy du
5 . (4)
dx du dx
For a composite function of two variables z ⫽ f (u, v), where u ⫽ g(x, y) and v ⫽ h(x, y), we would
naturally expect two formulas analogous to (4), since we can compute both ⭸z /⭸x and ⭸z /⭸y. The
Chain Rule for functions of two variables is summarized as follows:

Theorem 9.4.1 Chain Rule

If z ⫽ f (u, v) is differentiable and u ⫽ g(x, y) and v ⫽ h(x, y) have continuous first partial
derivatives, then
0z 0z 0u 0z 0v 0z 0z 0u 0z 0v
⫽ ⫹ , ⫽ ⫹ . (5)
0x 0u 0x 0v 0x 0y 0u 0y 0v 0y

498 | CHAPTER 9 Vector Calculus

 EXAMPLE 5 Chain Rule
If z  u  v and u  e2x3y, v  sin(x 2  y 2), find z/x and z/y.
2 3

SOLUTION Since z/u  2u and z/v  3v2, it follows from (5) that
0z
 2u(2e2x3y)  3v2 [2x cos(x 2  y2)]  4ue2x3y  6xv2 cos(x 2  y2) (6)
0x

0z
 2u(3e2x3y)  3v2 [(2y) cos(x2  y2)]  6ue2x3y  6yv2 cos(x2  y2). (7)
0y
Of course, in Example 5 we could substitute the expressions for u and v in the original func-
tion and then find the partial derivatives directly. In the same manner, the answers (6) and (7)
can be expressed in terms of x and y.

Special Case If z  f (u, v) is differentiable and u  g(t) and v  h(t) are differentiable
functions of a single variable t, then Theorem 9.4.1 implies that the ordinary derivative dz/dt is
z dz 0z du 0z dv
∂z ∂z 5 1 . (8)
∂u ∂v
dt 0u dt 0v dt
u v
Generalizations The results given in (5) and (8) immediately generalize to any number
∂u ∂u ∂v ∂v of variables. If z  f (u1, u2, . . . , un) and each of the variables u1, u2, u3, . . . , un are functions of
∂x ∂y ∂x ∂y
x1, x2, . . . , xk, then under the same assumptions as in Theorem 9.4.1 we have
x y x y 0z 0z 0u1 0z 0u2 0z 0un
(a) 5 1 1p 1 , (9)
0xi 0u1 0xi 0u2 0xi 0un 0xi
z
where i  1, 2, . . . , k. Similarly, if the ui, i  1, . . . , n, are differentiable functions of a single
∂z ∂z variable t, then
∂u ∂v
dz 0z du1 0z du2 0z dun
u v 5 1 1p 1 . (10)
dt 0u1 dt 0u2 dt 0un dt
du dv
dt dt
Tree Diagrams The results in (5) can be memorized in terms of a tree diagram. The dots
t t in the first diagram in FIGURE 9.4.5(a) indicate the fact that z depends on u and v; u and v depend, in
(b) turn, on x and y. To compute z /x for example, we read the diagram vertically downward starting
r from z and following the two blue polygonal paths leading to x, multiply the partial derivatives
on each path, and then add the products. The result given in (8) is represented by the second tree
diagram in Figure 9.4.5(b).
∂r ∂r ∂r
∂x ∂y ∂z
We shall use tree diagrams in the next two examples to illustrate special cases of (9) and (10).

x y z EXAMPLE 6 Using Tree Diagrams

∂x ∂x ∂y ∂y ∂z ∂z If r  x2  y5z3 and x  uve2s, y  u2  v2s, z  sin(uvs2), find r/s.
∂u ∂s ∂u ∂s ∂u ∂s
∂x ∂y ∂z SOLUTION From the blue and green paths in the tree diagram in Figure 9.4.5(c) we obtain
∂v ∂v ∂v
0r 0r 0x 0r 0y 0r 0z
u v s u v s u v s   
0s 0x 0s 0y 0s 0z 0s
(c)
z
 2x (2uve 2s)  5y 4z 3(v 2)  3y 5z 2(2uvs cos (uvs 2)).

EXAMPLE 7 Using Tree Diagrams

∂z ∂z ∂z
∂u ∂v ∂w If z  u2v3w4 and u  t 2, v  5t  8, w  t 3  t, find dz/dt.
u v w SOLUTION In this case the tree diagram in Figure 9.4.5(d) indicates that
du dv dw
dt dt dt dz 0z du 0z dv 0z dw
5 1 1
dt 0u dt 0v dt 0w dt
t t t
(d)  2uv3w4 (2t)  3u2v2w4 (5)  4u2v3w3 (3t 2  1).

FIGURE 9.4.5 Tree diagrams ALTERNATIVE SOLUTION Differentiate z  t 4 (5t  8)3 (t 3  t)4 by the Product Rule.

9.4 Partial Derivatives | 499

 REMARKS
If w  F(x, y, z) has continuous partial derivatives of any order, then analogous to (3), the
mixed partial derivatives are equal:

Fxyz  Fyzx  Fzyx, Fxxy  Fyxx  Fxyx ,

and so on.

9.4 Exercises Answers to selected odd-numbered problems begin on page ANS-22.

In Problems 1–6, sketch some of the level curves associated In Problems 33 and 34, verify that the given function satisfies
with the given function. Laplace’s equation:
1. f (x, y)  x  2y 2. f (x, y)  y2  x 0 2z 0 2z
3. f (x, y)  "x 2 y 2 12 2   0.
0x 2 0y 2
4. f (x, y)  "36 2 4x 2 2 9y 2 2
2 y2
33. z  ln(x2  y2) 34. z  e x cos 2xy
y 2 x2 1
5. f (x, y)  e 6. f (x, y)  tan ( y  x)
In Problems 35 and 36 verify that the given function satisfies the
In Problems 7–10, describe the level surfaces but do not graph. wave equation:
x2 z2 0 2u 0 2u
7. F(x, y, z)   a2  .
9 4 0x 2 0t 2
8. F(x, y, z)  x 2  y2  z2
9. F(x, y, z)  x 2  3y2  6z2 35. u  cos at sin x 36. u  cos(x  at)  sin(x  at)
10. F(x, y, z)  4y  2z  1 37. The molecular concentration C(x, t) of a liquid is given by
2

11. Graph some of the level surfaces associated with F(x, y, z)  C(x, t)  t 1/2 e x >kt . Verify that this function satisfies the
x 2  y2  z2 for c  0, c  0, and c  0. diffusion equation:
12. Given that k 0 2C 0C
2
 .
x2 y2 z2 4 0x 0t
F(x, y, z)    ,
16 4 9 38. The pressure P exerted by an enclosed ideal gas is given by
P  k(T/V), where k is a constant, T is temperature, and V is
find the x-, y-, and z-intercepts of the level surface that passes volume. Find:
through (4, 2, 3). (a) the rate of change of P with respect to V,
(b) the rate of change of V with respect to T, and
In Problems 13–32, find the first partial derivatives of the given (c) the rate of change of T with respect to P.
function.
13. z  x 2  xy2  4y5 14. z  x3  6x 2y3  5y2 In Problems 39–48, use the Chain Rule to find the indicated
4 3 2 6 5
15. z  5x y  x y  6x  4y partial derivatives.
16. z  tan(x 3y2) 2 0z 0z
39. z  e uv ; u  x 3, v  x  y2; ,
0x 0y
4"x 0z 0z
17. z  18. z  4x 3  5x 2  8x
3y 2  1 40. z  u2 cos 4v; u  x2y3, v  x 3  y3; ,
0x 0y
19. z  (x 3  y2)1 20. z  (x4  7y2  3y)6 0z 0z
21. z  cos2 5x  sin2 5y 22. z  e x
2
tan 1y2 41. z  4x  5y2; x  u4  8v3, y  (2u  v)2; ,
0u 0v
3 u x2y u v 2 0z 0z
23. f (x, y)  xe x y 24. f (u, f)  f2 sin 42. z  ; x , y ; ,
f xy v u 0u 0v
3x 2 y xy 0w 0w
25. f (x, y)  26. f (x, y)  43. w  (u2  v2)3/2; u  et sin u, v  et cos u; ,
x  2y (x 2 2 y 2)2 0t 0u
0w 0w
"r "s 44. w  tan1 !uv; u  r 2 2 s 2, v  r 2s 2; ,
27. g(u, v)  ln(4u2  5v3) 28. h(r, s)  2 0r 0s
s r
2 2 2 2 0R 0R
29. w  2 "xy  ye y/z
30. w  xy ln(xz) 45. R  rs2t 4; r  ue v , s  ve u , t  e u v ; ,
0u 0v
31. F(u, v, x, t)  u2w2  uv3  vw cos(ut 2)  (2x 2t)4 x x 0Q 0Q
4 5 46. Q  ln( pqr); p  t 2 sin1 x, q  2 , r  tan1 ; ,
32. G( p, q, r, s)  ( p2q 3)r s t t 0x 0t

500 | CHAPTER 9 Vector Calculus

47. w  "x 2  y 2; x  ln (rs  tu), temperature, respectively. If the equation defines V as a function
of P and T, and also defines T as a function of V and P, show that
t 0w 0w 0w 0F
y cosh rs; , ,
u 0t 0r 0u 0V 0T 1
48. s  p  q  r  4t; p  fe3u, q  cos(f  u), r  fu2,
2 2 2   .
0T 0F 0T
0s 0s
t  2f  8u; , 0V 0V
0f 0u 56. The voltage across a conductor is increasing at a rate of
In Problems 49–52, use (8) to find the indicated derivative. 2 volts/min and the resistance is decreasing at a rate of
dz 1 ohm/min. Use I  E/R and the Chain Rule to find the rate
49. z  ln(u2  v2); u  t 2, v  t 2; at which the current passing through the conductor is changing
dt
dz when R  50 ohms and E  60 volts.
3 4 5t
50. z  u v  uv ; u  e , v  sec 5t; 57. The length of the side labeled x of the triangle in FIGURE 9.4.6
dt
p p dw increases at a rate of 0.3 cm/s, the side labeled y increases
51. w  cos(3u  4v); u  2t  , v  t 2 ; 2 at a rate of 0.5 cm/s, and the included angle u increases at a
2 4 dt t p
4 dw rate of 0.1 rad/s. Use the Chain Rule to find the rate at which
52. w  e xy; x  , y  3t  5; 2 the area of the triangle is changing at the instant x  10 cm,
2t  1 dt t  0
y  8 cm, and u  p/6.
53. If u  f (x, y) and x  r cos u, y  r sin u, show that Laplace’s
equation 2u/x 2  2u/y2  0 becomes x
2 2
0u 1 0u 1 0u θ
2
  2 2  0.
0r r 0r r 0u y

54. Van der Waals’ equation of state for the real gas CO2 is FIGURE 9.4.6 Triangle in Problem 57
0.08T 3.6
P 2 2. 58. A particle moves in 3-space so that its coordinates at any time
V 2 0.0427 V are x  4 cos t, y  4 sin t, z  5t, t  0. Use the Chain Rule
If dT/dt and dV/dt are rates at which the temperature and vol- to find the rate at which its distance
ume change, respectively, use the Chain Rule to find dP/dt.
w  "x 2  y 2  z 2
55. The equation of state for a thermodynamic system is
F(P, V, T)  0, where P, V, and T are pressure, volume, and from the origin is changing at t  5p/2 seconds.

9.5 Directional Derivative

z INTRODUCTION We saw in the last section that for a function f of two variables x and y,
the partial derivatives z/x and z/y give the slope of the tangent to the trace, or curve of in-
tersection of the surface defined by z  f (x, y) and vertical planes that are, respectively, parallel
z = f(x, y) to the x- and y-coordinates axes. Equivalently, we can think of the partial derivative z/x as
the rate of change of the function f in the direction given by the vector i, and z/y as the rate of
change of the function f in the j-direction. There is no reason to confine our attention to just two
y
directions. In this section we shall see how to find the rate of change of a differentiable function
rate of change of f
in the j-direction in any direction. See FIGURE 9.5.1.
is ∂ z
x rate of change of f u ∂y The Gradient of a Function In this and the next sections it is convenient to introduce
in the i-direction What is the
a new vector based on partial differentiation. When the vector differential operator
is ∂ z rate of change of
∂x f in the direction
given by the vector u? 0 0 0 0 0
= 5i 1j or = 5 i 1j 1k
0x 0y 0x 0y 0z
FIGURE 9.5.1 An arbitrary direction is
denoted by the vector u is applied to a differentiable function z  f (x, y) or w  F(x, y, z), we say that the vectors
0f 0f
=f (x, y)  i j (1)
0x 0y

0F 0F 0F
=F(x, y, z)  i j k (2)
0x 0y 0z

9.5 Directional Derivative | 501

 are the gradients of the respective functions. The symbol , an inverted capital Greek delta, is
called “del” or “nabla.” The vector f is usually read “grad f.”

EXAMPLE 1 Gradient
Compute f (x, y) for f (x, y)  5y  x3y2.
0 0
SOLUTION From (1), f (x, y)  (5y  x3y2)i  (5y  x3y2) j; therefore
0x 0y

f (x, y)  3x2y2 i  (5  2x3y) j.

EXAMPLE 2 Gradient at a Point

If F(x, y, z)  xy  3x 2  z3, find F(x, y, z) at (2, 1, 4).
2

SOLUTION From (2), F(x, y, z)  ( y 2  6x) i  2xy j  3z2 k and so

F(2, 1, 4)  13 i  4 j  48 k.

A Generalization of Partial Differentiation Suppose u  cos u i  sin u j is a unit

vector in the xy-plane that makes an angle u with the positive x-axis and is parallel to the vector
v from (x, y, 0) to (x   x, y  y, 0). If h  "(Dx)2  (Dy)2  0, then v  h u. Furthermore,
let the plane perpendicular to the xy-plane that contains these points slice the surface z  f ( x, y)
in a curve C. We ask: What is the slope of the tangent line to C at a point P with coordinates
(x, y, f (x, y)) in the direction given by v? See FIGURE 9.5.2.
From the figure we see that x  h cos u and y  h sin u so that the slope of the indicated
secant line is
f (x  Dx, y  Dy) 2 f (x, y) f (x  h cos u, y  h sin u) 2 f (x, y)
 . (3)
h h
z

tangent
secant

surface
C z = f (x, y)
P
f(x + Δx, y + Δy) – f(x, y)

y
(x, y, 0)
θ u Δx v = hu
θ
Δy (x + Δx, y + Δy, 0)
x

FIGURE 9.5.2 C is the curve of intersection of the surface and the plane determined by vector v

We expect the slope of the tangent at P to be the limit of (3) as h S 0. This slope is the rate of change
of f at P in the direction specified by the unit vector u. This leads us to the following definition:

Definition 9.5.1 Directional Derivative

The directional derivative of z  f (x, y) in the direction of a unit vector u  cos u i  sin u j is
f (x  h cos u, y  h sin u) 2 f (x, y)
Du f (x, y)  lim (4)
hS0 h
provided the limit exists.

502 | CHAPTER 9 Vector Calculus

 Observe that (4) is truly a generalization of partial differentiation, since
f (x  h, y) 2 f (x, y) 0z
u0 implies that Di f (x, y)  lim 
hS0 h 0x

p f (x, y  h) 2 f (x, y) 0z
and u implies that Dj f (x, y)  lim  .
2 hS0 h 0y

Method for Computing the Directional Derivative While (4) could be used
to find Du f (x, y) for a given function, as usual we seek a more efficient procedure. The next
theorem will show how the concept of the gradient of a function plays a key role in computing
a directional derivative.

Theorem 9.5.1 Computing a Directional Derivative

If z  f (x, y) is a differentiable function of x and y and u  cos u i  sin u j, then

Du f (x, y)  f (x, y)  u. (5)

PROOF: Let x, y, and u be fixed so that g(t)  f (x  t cos u, y  t sin u) is a function of one
variable. We wish to compare the value of g (0), which is found by two different methods. First,
by the definition of a derivative,
g(0  h) 2 g(0) f (x  h cos u, y  h sin u) 2 f (x, y)
g9(0)  lim  lim . (6)
hS0 h hS0 h
Second, by the Chain Rule, (8) of Section 9.4,

d d
g9(t)  f1(x  t cos u, y  t sin u) (x  t cos u)  f2(x  t cos u, y  t sin u) (x  t sin u) (7)
dt dt
 f1(x  t cos u, y  t sin u) cos u  f2(x  t cos u, y  t sin u) sin u.

Here the subscripts 1 and 2 refer to the partial derivatives of f (x  t cos u, y  t sin u) with respect
to x  t cos u and y  t sin u, respectively. When t  0, we note that x  t cos u and y  t sin u
are simply x and y, and therefore (7) becomes
g (0)  fx(x, y) cos u  fy(x, y) sin u. (8)
Comparing (4), (6), and (8) then gives

Du f (x, y)  fx(x, y) cos u  fy(x, y) sin u

 [ fx(x, y) i  fy(x, y) j]  (cos u i  sin u j)

 f (x, y)  u.

EXAMPLE 3 Directional Derivative

Find the directional derivative of f (x, y)  2x2y3  6xy at (1, 1) in the direction of a unit vector
whose angle with the positive x-axis is p/6.
0f 0f
SOLUTION Since  4xy3  6y and  6x 2y2  6x, we have
0x 0y
f (x, y)  (4xy3  6y) i  (6x 2y2  6x) j and f (1, 1)  10 i  12 j.

"3 1
Now, at u  p/6, u  cos u i  sin u j becomes u  i  j. Therefore,
2 2
"3 1
Du f (1, 1)  =f (1, 1)  u  (10 i  12 j)  a i  jb  5"3  6.
2 2

9.5 Directional Derivative | 503

 EXAMPLE 4 Directional Derivative
Consider the plane that is perpendicular to the xy-plane and passes through the points P(2, 1)
and Q(3, 2). What is the slope of the tangent line to the curve of intersection of this plane with
the surface f (x, y)  4x2  y2 at (2, 1, 17) in the direction of Q?
!
SOLUTION We want Du f (2, 1) in the direction given by the vector PQ  i  j. But since
!
PQ is not a unit vector, we form u  (1/ "2) i  (1/ "2) j. Now,
f (x, y)  8x i  2y j and f (2, 1)  16 i  2 j.
Therefore, the required slope is

1 1
Du f (2, 1)  (16 i  2 j)  a i jb  9"2.
"2 "2

Functions of Three Variables For a function w  F(x, y, z) the directional derivative

is defined by
F(x  h cos a, y  h cos b, z  h cos g) 2 F(x, y, z)
DuF (x, y, z)  lim ,
hS0 h

where a, b, and g are the direction angles of the unit vector u measured relative to the positive
x-, y-, and z-axes, respectively.* But in the same manner as before, we can show that

Du F(x, y, z)  F(x, y, z)  u. (9)

Notice that since u is a unit vector, it follows from (10) of Section 7.3 that

Du f (x, y)  compuf (x, y) and Du F(x, y, z)  compuF(x, y, z).

In addition, (9) reveals that

0w
Dk F(x, y, z)  .
0z

EXAMPLE 5 Directional Derivative

Find the directional derivative of F(x, y, z)  xy2  4x 2y  z2 at (1, 1, 2) in the direction
of 6 i  2 j  3 k.
0F 0F 0F
SOLUTION We have  y2  8xy,  2xy  4x 2, and  2z so that
0x 0y 0z

F( x, y, z)  ( y2  8xy) i  (2xy  4x2) j  2z k

F(1, 1, 2)  9 i  6 j  4 k.

Since i 6i  2j  3k i  7 then u  67 i  27 j  37 k is a unit vector in the indicated direction.

It follows from (9) that

6 2 3 54
DuF(1, 1, 2)  (9 i  6 j  4 k)  a i 1 j 1 kb 5 .
7 7 7 7

Maximum Value of the Directional Derivative Let f represent a function of either

two or three variables. Since (5) and (9) express the directional derivative as a dot product, we
see from (5) of Theorem 7.3.2 that

Du f  i f i i u i cos f  i f i cos f, ( i u i  1),

*Note that the numerator in (4) can be written f (x  h cos a, y  h cos b) – f (x, y), where b  (p/2)  a.

504 | CHAPTER 9 Vector Calculus

 where f is the angle between f and u. Because 0  f  p, we have 1  cos f  1 and,
consequently, i f i  Du f  i f i . In other words:

The maximum value of the directional derivative is i f i and it occurs when u

(10)
has the same direction as f (when cos f  1).
The minimum value of the directional derivative is i f i and it occurs when u
(11)
and f have opposite directions (when cos f  1).

EXAMPLE 6 Max/Min of Directional Derivative

In Example 5 the maximum value of the directional derivative of F of (1, 1, 2) is
i F(1, 1, 2)i  "133. The minimum value of DuF(1, 1, 2) is then  "133.

Gradient Points in Direction of Most Rapid Increase of f Put yet another way,
(10) and (11) state:
The gradient vector f points in the direction in which f increases most rapidly,
whereas f points in the direction of the most rapid decrease of f.

EXAMPLE 7 Direction of Steepest Ascent

road Each year in Los Angeles there is a bicycle race up to the top of a hill by a road known to be
the steepest in the city. To understand why a bicyclist with a modicum of sanity will zigzag
up the road, let us suppose the graph of f (x, y)  4  23 "x 2  y 2 , 0  z  4, shown in
∇f u FIGURE 9.5.3(a) is a mathematical model of the hill. The gradient of f is

(a) (b) 2
2 x y 3
f (x, y)  c i jd  r,
FIGURE 9.5.3 Model of a hill in 3 "x 2  y 2 2
"x  y 2
"x  y 2
2

Example 7
where r  x i  y j is a vector pointing to the center of the circular base.
Thus the steepest ascent up the hill is a straight road whose projection in the xy-plane is a
radius of the circular base. Since Du f  compuf, a bicyclist will zigzag, or seek a direction
u other than f, in order to reduce this component.

EXAMPLE 8 Direction to Cool Off Fastest

The temperature in a rectangular box is approximated by

T(x, y, z)  xyz(1  x)(2  y)(3  z), 0  x  1, 0  y  2, 0  z  3.

If a mosquito is located at ( 12 , 1, 1), in which direction should it fly to cool off as rapidly as
possible?
SOLUTION The gradient of T is

T(x, y, z)  yz(2  y)(3  z)(1  2x)i  xz(1  x)(3  z)(2  2y)j  xy(1  x)(2  y)(3  2z)k.

Therefore, T ( 12 , 1, 1)  14 k. To cool off most rapidly, the mosquito should fly in the direction
of 14 k; that is, it should dive for the floor of the box, where the temperature is T(x, y, 0)  0.

9.5 Exercises Answers to selected odd-numbered problems begin on page ANS-22.

In Problems 1–4, compute the gradient for the given function. In Problems 5–8, find the gradient of the given function at the
1. f (x, y)  x 2  x 3y2  y4
2
2. f (x, y)  y  e22x y indicated point.
xy2 5. f (x, y)  x 2  4y2; (2, 4)
3. F(x, y, z)  4. F(x, y, z)  xy cos yz
z3 6. f (x, y)  "x 3y 2 y 4; (3, 2)

9.5 Directional Derivative | 505

 7. F(x, y, z)  x 2z2 sin 4y; (2, p/3, 1) 34. Suppose Du f (a, b)  6. What is the value of Du f (a, b)?
8. F(x, y, z)  ln(x 2  y2  z2); (4, 3, 5) 35. (a) If f (x, y)  x3  3x2y2  y3, find the directional derivative
In Problems 9 and 10, use Definition 9.5.1 to find Du f (x, y) of f at a point (x, y) in the direction of u  (1/ "10)(3i  j).
given that u makes the indicated angle with the positive x-axis. (b) If F(x, y)  Du f (x, y) in part (a), find DuF(x, y).
36. Consider the gravitational potential
9. f (x, y)  x 2  y2; u  30
10. f (x, y)  3x  y2; u  45 Gm
U(x, y)  ,
In Problems 11–20, find the directional derivative of the given "x 2  y 2
function at the given point in the indicated direction.
11. f (x, y)  5x3y6; (1, 1), u  p/6
where G and m are constants. Show that U increases or
12. f (x, y)  4x  xy2  5y; (3, 1), u  p/4
decreases most rapidly along a line through the origin.
y 37. If f (x, y)  x 3  12x  y2  10y, find all points at which
13. f (x, y)  tan1 ; (2, 2), i  3 j i f i  0.
x
xy 38. Suppose
14. f (x, y)  ; (2, 1), 6 i  8 j
xy Du f (a, b)  7, Dv f (a, b)  3
15. f (x, y)  (xy  1)2; (3, 2), in the direction of (5, 3)
16. f (x, y)  x2 tan y; (12, p>3), in the direction of the negative x-axis 5 12 5 12
17. F(x, y, z)  x 2y2 (2z  1)2; (1, 1, 1), 0, 3, 3 u i2 j, v  i j.
13 13 13 13
x 2 2 y2
18. F(x, y, z)  ; (2, 4, 1), i  2 j  k Find f (a, b).
z2 39. Consider the rectangular plate shown in FIGURE 9.5.4. The tem-
19. F(x, y, z)  "x 2y  2y 2z; (2, 2, 1), in the direction of the perature at a point (x, y) on the plate is given by T(x, y) 
negative z-axis 5  2x 2  y2. Determine the direction an insect should take,
20. F(x, y, z)  2x  y2  z2; (4, 4, 2), in the direction of the starting at (4, 2), in order to cool off as rapidly as possible.
origin
In Problems 21 and 22, consider the plane through the points P y
and Q that is perpendicular to the xy-plane. Find the slope of the (4, 2)
tangent at the indicated point to the curve of intersection of this
plane and the graph of the given function in the direction of Q.
x
21. f (x, y)  (x  y)2; P(4, 2), Q(0, 1); (4, 2, 4)
22. f (x, y)  x3  5xy  y2; P(1, 1), Q(1, 6); (1, 1, 3)
In Problems 23–26, find a vector that gives the direction in
which the given function increases most rapidly at the indicated
point. Find the maximum rate.
FIGURE 9.5.4 Insect in Problem 39
23. f (x, y)  e2x sin y; (0, p/4)
24. f (x, y)  xyexy; (5, 5)
25. F(x, y, z)  x 2  4xz  2yz2; (1, 2, 1) 40. In Problem 39, observe that (0, 0) is the coolest point of the
26. F(x, y, z)  xyz; (3, 1, 5) plate. Find the path the cold-seeking insect, starting at (4, 2),
will take to the origin. If x(t), y(t) is the vector equation of
In Problems 27–30, find a vector that gives the direction in the path, then use the fact that T(x, y)  x (t), y (t). Why
which the given function decreases most rapidly at the indicated is this? [Hint: Remember separation of variables?]
point. Find the minimum rate. 41. The temperature at a point (x, y) on a rectangular metal plate
27. f (x, y)  tan(x 2  y2); ( "p>6, "p>6) is given by T(x, y)  100  2x 2  y2. Find the path a heat-
28. f (x, y)  x 3  y3; (2, 2) seeking particle will take, starting at (3, 4), as it moves in the
29. F(x, y, z)  "xzey; (16, 0, 9) direction in which the temperature increases most rapidly.
42. The temperature T at a point (x, y, z) in space is inversely
xy
30. F(x, y, z)  ln a b; (12, 16, 13) proportional to the square of the distance from (x, y, z) to
z
the origin. It is known that T(0, 0, 1)  500. Find the rate of
31. Find the directional derivative(s) of f (x, y)  x  y 2 at
change of T at (2, 3, 3) in the direction of (3, 1, 1). In which
(3, 4) in the direction of a tangent vector to the graph of
direction from (2, 3, 3) does the temperature T increase most
2x 2  y2  9 at (2, 1).
rapidly? At (2, 3, 3) what is the maximum rate of change of T?
32. If f (x, y)  x2  xy  y2  x, find all points where Du f (x, y)
43. Find a function f such that
in the direction of u  (1/ "2)(i  j) is zero.
33. Suppose f (a, b)  4i  3j. Find a unit vector u so that f  (3x2  y3  ye xy) i  (2y2  3xy2  xe xy) j.
(a) Du f (a, b)  0,
(b) Du f (a, b) is a maximum, and 44. Let fx, fy, fxy, fyx be continuous and u and v be unit vectors.
(c) Du f (a, b) is a minimum. Show that DuDv f  DvDu f.

506 | CHAPTER 9 Vector Calculus

In Problems 45–48, assume that f and g are differentiable 49. If F(x, y, z)  f1(x, y, z) i  f2(x, y, z) j  f3(x, y, z) k and
functions of two variables. Prove the given identity. 0 0 0
= 5i 1j 1k ,
45. (cf )  c f 46. ( f  g)  f  g 0x 0y 0z
f g=f 2 f =g show that
47. ( fg)  f g  gf 48. =a b 5
g g2 0f3 0f2 0f1 0f3 0f2 0f1
= 3F5 a 2 bi1 a 2 bj1 a 2 b k.
0y 0z 0z 0x 0x 0y

9.6 Tangent Planes and Normal Lines

INTRODUCTION The notion of the gradient of a function of two or more variables was
introduced in the preceding section as an aid in computing directional derivatives. In this section
we give a geometric interpretation of the gradient vector.

Geometric Interpretation of the Gradient—Functions of Two Variables

Suppose f (x, y)  c is the level curve of the differentiable function z  f (x, y) that passes through
a specified point P(x0, y0); that is, f (x0, y0)  c. If this level curve is parameterized by the
differentiable functions
curve
f(x, y) = c x  g(t), y  h(t) such that x0  g(t0), y0  h(t0),

r′(t0) then the derivative of f (g(t), h(t))  c with respect to t is

0f dx 0f dy
  0. (1)
0x dt 0y dt
P(x0, y0) ∇f (x0, y0)
When we introduce the vectors
0f 0f dx dy
f (x, y)  i j and r (t)  i j,
0x 0y dt dt

FIGURE 9.6.1 Gradient is perpendicular (1) becomes f  r  0. Specifically, at t  t0, we have

to tangent vector at P f (x0, y0)  r (t0)  0. (2)
Thus, if r (t0) 0, the vector f (x0, y0) is orthogonal to the tangent vector r (t0) at P(x0, y0). We
interpret this to mean:
∇f(2, 3)
f is orthogonal to the level curve at P.
See FIGURE 9.6.1.
y

(2, 3)
EXAMPLE 1 Gradient at a Point
x
Find the level curve of f (x, y)  x 2  y2 passing through (2, 3). Graph the gradient at the point.
–x2 + y2 = 5 SOLUTION Since f (2, 3)  4  9  5, the level curve is the hyperbola x 2  y2  5. Now,
f (x, y)  2xi  2yj and f (2, 3)  4 i  6 j.
FIGURE 9.6.2 Gradient in Example 1 FIGURE 9.6.2 shows the level curve and f (2, 3).

Geometric Interpretation of the Gradient—Functions of Three Variables

Proceeding as before, let F(x, y, z)  c be the level surface of a differentiable function w  F(x, y, z)
that passes through P(x0, y0, z0). If the differentiable functions x  f (t), y  g(t), z  h(t) are the
parametric equations of a curve C on the surface for which x0  f (t0), y0  g(t0), z0  h(t0), then
the derivative of F( f (t), g(t), h(t))  0 implies that
0F dx 0F dy 0F dz
  0
0x dt 0y dt 0z dt

9.6 Tangent Planes and Normal Lines | 507

 ∇F(x0, y0, z0) 0F 0F 0F dx dy dz
or a i j kb  a i  j kb  0. (3)
0x 0y 0z dt dt dt

In particular, at t  t0, (3) is

C
F(x0, y0, z0)  r (t0)  0. (4)
r′(t0)
P(x0, y0, z0)
Thus, when r (t0) 0, the vector F(x0, y0, z0) is orthogonal to the tangent vector r (t0). Since this
surface argument holds for any differentiable curve through P(x0, y0, z0) on the surface, we conclude that:
F(x, y, z) = c
F is normal ( perpendicular) to the level surface at P.
FIGURE 9.6.3 Gradient is normal to level
surface at P See FIGURE 9.6.3.

z EXAMPLE 2 Gradient at a Point

Find the level surface of F(x, y, z)  x 2  y2  z2 passing through (1, 1, 1). Graph the
gradient at the point.
∇F (1, 1, 1)
SOLUTION Since F(1, 1, 1)  3, the level surface passing through (1, 1, 1) is the sphere
(1, 1, 1) x2  y2  z2  3. The gradient of the function is
y

F(x, y, z)  2x i  2y j  2z k,
x2 + y2 + z2 = 3

x and so, at the given point, F(1, 1, 1)  2 i  2 j  2 k. The level surface and F(1, 1, 1) are
illustrated in FIGURE 9.6.4.

FIGURE 9.6.4 Gradient in Example 2 Tangent Plane In the study of differential calculus a basic problem was finding an equa-
tion of a tangent line to the graph of a function. In 3-space the analogous problem is finding an
equation of a tangent plane to a surface. We assume again that w  F(x, y, z) is a differentiable
function and that a surface is given by F(x, y, z)  c.

z tangent
plane at Definition 9.6.1 Tangent Plane
(x0, y0, z0)
Let P(x0, y0, z0) be a point on the graph of F(x, y, z)  c, where F is not 0. The tangent plane
∇F(x0, y0, z0) at P is that plane through P that is normal to F evaluated at P.

(x0, y0, z0)

surface
Thus, if P(x, y, z) and P(x0, y0, z0) are points on the tangent plane and r and r0 are their respective
F(x, y, z) = c position vectors, then a vector equation of the tangent plane is F(x0, y0, z0)  (r  r0)  0. See
y FIGURE 9.6.5. We summarize this last result:
(x, y, z)

x Theorem 9.6.1 Equation of Tangent Plane

FIGURE 9.6.5 Tangent plane is normal to Let P(x0, y0, z0) be a point on the graph of F(x, y, z)  c, where F is not 0. Then an equation
gradient at P of the tangent plane at P is

Fx(x0, y0, z0)(x  x0)  Fy(x0, y0, z0)( y  y0)  Fz(x0, y0, z0)(z  z0)  0. (5)

EXAMPLE 3 Equation of Tangent Plane

Find an equation of the tangent plane to the graph of x 2  4y2  z 2  16 at (2, 1, 4).
SOLUTION By defining F(x, y, z)  x2  4y2  z2, the given surface is the level surface
F(x, y, z)  F(2, 1, 4)  16 passing through (2, 1, 4). Now, Fx(x, y, z)  2x, Fy(x, y, z)  8y,
and Fz(x, y, z)  2z, so that

F(x, y, z)  2x i  8y j  2z k and F(2, 1, 4)  4 i  8 j  8 k.

508 | CHAPTER 9 Vector Calculus

 It follows from (5) that an equation of the tangent plane is

4(x  2)  8( y  1)  8(z  4)  0 or x  2y  2z  8.

Surfaces Given by z ⴝ f (x, y ) For a surface given explicitly by a differentiable

function z  f (x, y), we define F(x, y, z)  f (x, y)  z or F(x, y, z)  z  f (x, y). Thus, a point
(x0, y0, z0) is on the graph of z  f (x, y) if and only if it is also on the level surface F(x, y, z)  0.
This follows from F(x0, y0, z0)  f (x0, y0)  z0  0.

EXAMPLE 4 Equation of Tangent Plane

Find an equation of the tangent plane to the graph of z  12 x2  12 y2  4 at (1, 1, 5).
z
SOLUTION Define F(x, y, z)  12 x2  12 y2  z  4 so that the level surface of F passing
5 through the given point is F(x, y, z)  F(1, 1, 5) or F(x, y, z)  0. Now, Fx  x, Fy  y,
and Fz  1, so that

∇F(1, –1, 5) F(x, y, z)  x i  y j  k and F(1, 1, 5)  i  j  k.

y Thus, from (5) the desired equation is

(1, –1, 0)

x (x  1)  ( y  1)  (z  5)  0 or x  y  z  7.

FIGURE 9.6.6 Tangent plane in See FIGURE 9.6.6.

Example 4
Normal Line Let P(x0, y0, z0) be a point on the graph of F(x, y, z)  c, where F is
not 0. The line containing P(x0, y0, z0) that is parallel to F(x0, y0, z0) is called the normal
line to the surface at P. As indicated by its name, this line is normal to the tangent plane to
the surface at P.

EXAMPLE 5 Normal Line to a Surface

Find parametric equations for the normal line to the surface in Example 4 at (1, 1, 5).
SOLUTION A direction vector for the normal line at (1, 1, 5) is F(1, 1, 5)  i  j  k.
It follows that parametric equations for the normal line are

x  1  t, y  1  t, z  5  t.

Expressed as symmetric equations the normal line to a surface F (x, y, z) = c at P(x0, y0, z0)
is given by

x 2 x0 y 2 y0 z 2 z0
  .
Fx(x0, y0, z0) Fy(x0, y0, z0) Fz(x0, y0, z0)

In Example 5, you should verify that symmetric equations of the normal line at (1, 1, 5)
are
stream y11 z25
x215 5 .
21 21
100
80 P
60
40 REMARKS
30
Water flowing down a hill chooses a path in the direction of the greatest change in altitude.
contours FIGURE 9.6.7 shows the contours, or level curves, of a hill. As shown in the figure, a stream start-
of a hill
ing at point P will take a path that is perpendicular to the contours. After reading Sections 9.5
FIGURE 9.6.7 Stream is perpendicular to and 9.6, the student should be able to explain why.
contours

9.6 Tangent Planes and Normal Lines | 509

 9.6 Exercises Answers to selected odd-numbered problems begin on page ANS-22.
In Problems 1–12, sketch the level curve or surface passing 26. x 2  2y2  3z2  33; 8x  4y  6z  5
through the indicated point. Sketch the gradient at the point. 27. Find points on the surface x 2  4x  y2  z2  2z  11
y  2x at which the tangent plane is horizontal.
1. f (x, y)  x  2y; (6, 1) 2. f (x, y)  ; (1, 3)
x 28. Find points on the surface x 2  3y2  4z2  2xy  16 at
3. f (x, y)  y  x 2; (2, 5) 4. f (x, y)  x 2  y2; (1, 3) which the tangent plane is parallel to (a) the xz-plane, (b) the
x 2
y 2 yz-plane, and (c) the xy-plane.
5. f (x, y)   ; (2, 3)
4 9 In Problems 29 and 30, show that the second equation is an
y2 equation of the tangent plane to the graph of the first equation at
6. f (x, y)  ; (2, 2)
x (x0, y0, z0).
7. f (x, y)  (x  1)2  y2; (1, 1) x2 y2 z2 xx0 yy0 zz0
y21 29. 2  2  2  1; 2  2  2  1
8. f (x, y)  ; (p/6, 32) a b c a b c
sin x x 2
y 2
z 2
xx yy zz
0 0 0
9. F(x, y, z)  y  z; (3, 1, 1) 30. 2 2 2  2  1; 2 2 2  2  1
10. F(x, y, z)  x 2  y2  z; (1, 1, 3)
a b c a b c
31. Show that every tangent plane to the graph of z2  x2  y2
11. F(x, y, z)  "x 2  y 2  z 2 ; (3, 4, 0) passes through the origin.
12. F(x, y, z)  x 2  y2  z; (0, 1, 1) 32. Show that the sum of the x-, y-, and z-intercepts of every
In Problems 13 and 14, find the points on the given surface at tangent plane to the graph of "x  "y  "z  "a,
which the gradient is parallel to the indicated vector. a  0, is the number a.
13. z  x 2  y2; 4 i  j  12 k
In Problems 33 and 34, find parametric equations for the normal
14. x 3  y2  z  15; 27 i  8 j  k
line at the indicated point. In Problems 35 and 36, find symmet-
In Problems 15–24, find an equation of the tangent plane to the ric equations for the normal line.
graph of the given equation at the indicated point. 33. x 2  2y2  z2  4; (1, 1, 1)
15. x 2  y2  z2  9; (2, 2, 1) 34. z  2x 2  4y2; (3, 2, 2)
16. 5x 2  y2  4z2  8; (2, 4, 1) 35. z  4x 2  9y2  1; ( 12 , 13 , 3)
17. x 2  y2  3z2  5; (6, 2, 3) 36. x 2  y2  z2  0; (3, 4, 5)
18. xy  yz  zx  7; (1, 3, 5) 37. Show that every normal line to the graph x 2  y2  z2  a2
19. z  25  x 2  y2; (3, 4, 0) passes through the origin.
20. xz  6; (2, 0, 3) 38. Two surfaces are said to be orthogonal at a point P of inter-
21. z  cos(2x  y); (p/2, p/4, 1/ "2) section if their normal lines at P are orthogonal. Prove that
22. x2y3  6z  10; (2, 1, 1) the surfaces given by F(x, y, z)  0 and G(x, y, z)  0 are
orthogonal at P if and only if FxGx  FyGy  FzGz  0.
23. z  ln(x 2  y2); (1/ "2, 1/ "2, 0)
24. z  8e2y sin 4x; (p/24, 0, 4) In Problems 39 and 40, use the result of Problem 38 to show that
the given surfaces are orthogonal at a point of intersection.
In Problems 25 and 26, find the points on the given surface at
which the tangent plane is parallel to the indicated plane. 39. x 2  y2  z2  25; x 2  y2  z2  0
40. x 2  y2  z2  4; z  1/xy2
25. x 2  y2  z2  7; 2x  4y  6z  1

9.7 Curl and Divergence

INTRODUCTION In Section 9.1 we introduced the concept of vector function of one variable.
In this section we examine vector functions of two and three variables.

Vector Fields Vector functions of two and three variables,

F(x, y)  P(x, y) i  Q(x, y) j

F(x, y, z)  P(x, y, z) i  Q(x, y, z) j  R(x, y, z) k

510 | CHAPTER 9 Vector Calculus

 are also called vector fields. For example, the motion of a wind or a fluid can be described by
means of a velocity field because a vector can be assigned at each point representing the veloc-
ity of a particle at the point. See FIGURE 9.7.1(a) and 9.7.1(b). The concept of a force field plays
an important role in mechanics, electricity, and magnetism. See Figure 9.7.1(c) and 9.7.1(d).

+ +
va

vb

(a) Airflow around an airplane (b) Laminar flow of blood (c) Inverse square force (d) Lines of force around
wing in an artery; cylindrical field; magnitude two equal positive charges
layers of blood flow of the attractive
faster near the center force is large near
of the artery the particle

FIGURE 9.7.1 Various vector fields

EXAMPLE 1 Two-Dimensional Vector Field

Graph the two-dimensional vector field F(x, y)  y i  x j.
SOLUTION One manner of proceeding is simply to choose points in the xy-plane and then graph
the vector F at that point. For example, at (1, 1) we would draw the vector F(1, 1)  i  j.
For the given vector field it is possible to systematically draw vectors of the same length.
Observe that i Fi  "x 2  y 2 , and so vectors of the same length k must lie along the curve
defined by "x 2  y 2  k; that is, at any point on the circle x 2  y2  k2 a vector would
have length k. For simplicity let us choose circles that have some points on them with integer
coordinates. For example, for k  1, k  "2, and k  2, we have:
y

x2  y2  1: At the points (1, 0), (0, 1), (1, 0), (0, 1) the corresponding vectors
j, i, j, i have the same length 1.

x x2  y2  2: At the points (1, 1), (1, 1), (1, 1), (1, 1) the corresponding
1 2
vectors i  j, i  j, i  j, i  j have the same length "2.

x2  y2  4: At the points (2, 0), (0, 2), (2, 0), (0, 2) the corresponding vectors
2 j, 2 i, 2 j, 2 i have the same length 2.

FIGURE 9.7.2 Vector field in Example 1 The vectors at these points are shown in FIGURE 9.7.2.

In Section 9.5 we saw that the del operator

0 0 0
= i j k
0x 0y 0z

combined with a scalar function f(x, y, z) produces a vector field

0f 0f 0f
F(x, y, z)  =f  i j k
0x 0y 0z

called the gradient of f or a gradient field. The del operator can also be combined with a vector
field F(x, y, z)  P(x, y, z) i  Q(x, y, z) j  R(x, y, z) k in two different ways: in one case produc-
ing another vector field and in the other producing a scalar function. We will assume hereafter
that P, Q, and R have continuous partial derivatives.

9.7 Curl and Divergence | 511

 Definition 9.7.1 Curl
The curl of a vector field F  P i  Q j  R k is the vector field

0R 0Q 0P 0R 0Q 0P
curl F  a 2 bi a 2 bj a 2 b k.
0y 0z 0z 0x 0x 0y

In practice, curl F can be computed from the cross product of the del operator and the
vector F:

i j k
0 0 0
curl F   F 4 4. (1)
0x 0y 0z
P Q R

n There is another combination of partial derivatives of the component functions of a vector field
that occurs frequently in science and engineering. Before we state the next definition, consider
the following motivation.
F
compnF If F(x, y, z)  P(x, y, z) i  Q(x, y, z) j  R(x, y, z) k is the velocity field of a fluid, then as
shown in FIGURE 9.7.3, the volume of the fluid flowing through an element of surface area S
per unit time—that is, the flux of the vector field F through the area S—is approximated by

ΔS (height)(area of base)  (compn F) S  (F  n) S, (2)

FIGURE 9.7.3 Fluid flow through element
where n is a unit vector normal to the surface. Now consider the rectangular parallelepiped shown
of area S
in FIGURE 9.7.4. To compute the total flux of F through its six sides in the outward direction we
first compute the total flux out of parallel faces. The area of face F1 is x z and its outward unit
z normal is j, and so by (2) the flux of F through F1 is approximately

Δz F  (j) x z  Q(x, y, z) x z.
F1 F2
The flux out of face F2, whose outward normal is j, is approximated by
Δx
(x, y, z)
Δy (F  j) x z  Q(x, y  y, z) x z.
y
Consequently the total flux out of these parallel faces is
x

FIGURE 9.7.4 What is total flux of vector Q(x, y  y, z) x z  (Q(x, y, z) x z)  [Q(x, y  y, z)  Q(x, y, z)] x z. (3)
field through this element?
By multiplying (3) by y/ y and recalling the definition of a partial derivative, we get for
y close to 0,

Q(x, y  Dy, z) 2 Q(x, y, z) 0Q

Dx Dy Dz < Dx Dy Dz.
Dy 0y

Arguing in exactly the same manner, we see that the contributions to the total flux out of the
parallelepiped from the two faces parallel to the yz-plane, and from the two faces parallel to the
xy-plane, are, in turn,

0P 0R
Dx Dy Dz and Dx Dy Dz.
0x 0z

Adding the results, we see that the net flux of F out of the parallelepiped is approximately

0P 0Q 0R
a   b Dx Dy Dz.
0x 0y 0z

512 | CHAPTER 9 Vector Calculus

By dividing the last expression by x y z, we get the outward flux of F per unit volume:

0P 0Q 0R
1 1 .
0x 0y 0z

It is this combination of partial derivatives that is given a special name.

Definition 9.7.2 Divergence

The divergence of a vector field F  P i  Q j  R k is the scalar function

0P 0Q 0R
div F    .
0x 0y 0z

Observe that div F can also be written in terms of the del operator as

0 0 0
div F    F  P(x, y, z)  Q(x, y, z)  R(x, y, z). (4)
0x 0y 0z

EXAMPLE 2 Curl and Divergence

If F  (x 2y3  z4) i  4x 5y2z j  y4z6 k, find (a) curl F, (b) div F, and (c) div(curl F).
SOLUTION (a) From (1),

i j k
0 0 0
curl F   F 4 4
0x 0y 0z
x2y3 2 z4 4x5y2z 2y4z6

 (4y3z6  4x5y2) i  4z3 j  (20x4y2z  3x2y2) k.

(b) From (4),

0 0 0
div F    F  (x 2y3  z4)  (4x 5y2z)  (y4z6)
0x 0y 0z

 2xy3  8x 5yz  6y4z5.

(c) From Definition 9.7.2 and part (a) we find

0 0 0
div(curl F)  (4y3z6  4x5y2)  (4z3)  (20x4y2z  3x2y2)
0x 0y 0z

 0  20x4y2  0  20x 4y2  0.

We ask you to prove the following two important properties. If f is a scalar function with
continuous second partial derivatives, then

curl(grad f )   f  0. (5)

Also, if F is a vector field having continuous second partial derivatives, then

div(curl F)    ( F)  0. (6)

See part (c) of Example 2 and Problems 29 and 30 in Exercises 9.7.

9.7 Curl and Divergence | 513

 w Physical Interpretations The word curl was introduced by James Clerk Maxwell
A
(1831–1879), a Scottish physicist, in his studies of electromagnetic fields. However, the curl
is easily understood in connection with the flow of fluids. If a paddle device, such as shown
B in FIGURE 9.7.5, is inserted in a flowing fluid, then the curl of the velocity field F is a measure
of the tendency of the fluid to turn the device about its vertical axis w. If curl F  0, then the
flow of the fluid is said to be irrotational, which means that it is free of vortices or whirlpools
that would cause the paddle to rotate.* In FIGURE 9.7.6 the axis w of the paddle points straight
out of the page.
FIGURE 9.7.5 Paddle device

A A A A B
B B B B B A A

(a) Irrotational flow (b) Rotational flow

FIGURE 9.7.6 Irrotational flow in (a); rotational flow in (b)

P

(a) div F(P) > 0; P a source

In the motivational discussion leading to Definition 9.7.2 we saw that the divergence of a
velocity field F near a point P(x, y, z) is the flux per unit volume. If div F(P)  0, then P is said
to be a source for F, since there is a net outward flow of fluid near P; if div F(P)  0, then P is
P said to be a sink for F, since there is a net inward flow of fluid near P; if div F(P)  0, there are
no sources or sinks near P. See FIGURE 9.7.7.
(b) div F(P) < 0; P a sink The divergence of a vector field can also be interpreted as a measure of the rate of change of
the density of the fluid at a point. In other words, div F is a measure of the fluid’s compressibility.
FIGURE 9.7.7 P a source in (a); P a sink If   F  0, the fluid is said to be incompressible. In electromagnetic theory, if   F  0, the
in (b) vector field F is said to be solenoidal.

*In science texts the word rotation is sometimes used instead of curl. The symbol curl F is then replaced
by rot F.

9.7 Exercises Answers to selected odd-numbered problems begin on page ANS-23.

In Problems 1–6, graph some representative vectors in the given 12. F(x, y, z)  5y3 i  ( 12 x3y2  xy) j  (x3yz  xz) k
vector field. 13. F(x, y, z)  xez i  4yz2 j  3yez k
1. F(x, y)  x i  y j 14. F(x, y, z)  yz ln x i  (2x  3yz) j  xy2z3 k
2. F(x, y)  x i  y j 15. F(x, y, z)  xye x i  x3yzez j  xy2 ey k
3. F(x, y)  y i  x j 16. F(x, y, z)  x 2 sin yz i  z cos xz3 j  ye5xy k
4. F(x, y)  x i  2y j
5. F(x, y)  y j In Problems 17–24, let a be a constant vector and r  x i  y j  z k.
6. F(x, y)  x j
Verify the given identity.
17. div r  3
In Problems 7–16, find the curl and the divergence of the given
18. curl r  0
vector field.
19. (a ) r  2 a
7. F(x, y, z)  xz i  yz j  xy k 20.  (a r)  2 a
8. F(x, y, z)  10yz i  2x 2z j  6x3 k 21.   (a r)  0
9. F(x, y, z)  4xy i  (2x 2  2yz) j  (3z2  y2) k 22. a ( r)  0
10. F(x, y, z)  (x  y)3 i  eyz j  xye2y k 23.  [(r  r) a]  2(r a)
11. F(x, y, z)  3x 2y i  2xz3 j  y4 k 24.   [(r  r) a]  2(r  a)

514 | CHAPTER 9 Vector Calculus

In Problems 25–32, verify the given identity. Assume continuity
of all partial derivatives. v
P
25.   (F  G)    F    G
26.  (F  G)   F   G
27.   ( f F)  f (  F)  F  f
28.  ( f F)  f ( F)  (f ) F r
29. curl(grad f )  0 axis
30. div(curl F)  0
31. div(F G)  G  curl F  F  curl G
32. curl(curl F  grad f )  curl(curl F) O
33. Show that
FIGURE 9.7.8 Rotating body in Problem 40

0 2f 0 2f 0 2f In Problems 41 and 42, assume that f and g have continuous

  f  1 1 .
0x2 0y2 0z2 second partial derivatives. Show that the given vector field is
solenoidal. [Hint: See Problem 31.]
This is known as the Laplacian and is also written 2f. 41. F  f g
34. Show that   ( f f )  f 2 f  i f i 2 , where 2 f is the 42. F  f ( f g)
Laplacian defined in Problem 33. [Hint: See Problem 27.] 43. The velocity vector field for the two-dimensional flow of an
35. Find curl(curl F) for the vector field F  xy i  4yz2 j  ideal fluid around a cylinder is given by
2xz k.
36. (a) Assuming continuity of all partial derivatives, show that
x 2 2 y2 2xy
F(x, y)  A c a1 2 2 2 2
bi2 2 jd
curl(curl F)  2 F  grad(div F), where (x  y ) (x  y 2)2
for some positive constant A. See FIGURE 9.7.9.
2F  2 (P i  Q j  R k)  2P i  2Q j  2 R k. (a) Show that when the point (x, y) is far from the origin,
F(x, y)  A i.
(b) Show that F is irrotational.
(b) Use the identity in part (a) to obtain the result in
(c) Show that F is incompressible.
Problem 35.
37. Any scalar function f for which 2f  0 is said to be harmonic. y
Verify that f (x, y, z)  (x2  y2  z2)1/2 is harmonic except
at the origin. 2f  0 is called Laplace’s equation.
38. Verify that
x

2
f (x, y)  arctan a 2 2
b, x2  y2 1
x y 21

satisfies Laplace’s equation in two variables

FIGURE 9.7.9 Vector field in Problem 43

0 2f 0 2f
2 f  1  0. 44. If E  E(x, y, z, t) and H  H(x, y, z, t) represent electric and
0x 2 0y2 magnetic fields in empty space, then Maxwell’s equations
are
39. Let r  x i  y j  z k be the position vector of a mass m1
1 0H
and let the mass m2 be located at the origin. If the force of div E 5 0, curl E 5 2 ,
c 0t
gravitational attraction is
1 0E
Gm1m2 div H 5 0, curl H 5 ,
c 0t
F52 r,
iri 3 where c is the speed of light. Use the identity in Problem 36(a)
to show that E and H satisfy
verify that curl F  0 and div F  0, r 0.
40. Suppose a body rotates with a constant angular velocity ␻ 1 02 E 1 02 H
about an axis. If r is the position vector of a point P on the body =2 E 5 , =2 H 5 .
c2 0t2 c2 0t2
measured from the origin, then the linear velocity vector v of
rotation is v  ␻ r. See FIGURE 9.7.8. If r  x i  y j  z k 45. Consider the vector field F  x2yz i  xy2z j  (z  5x) k.
and ␻  v1 i  v2 j  v3 k, show that ␻  12 curl v. Explain why F is not the curl of another vector field G.

9.7 Curl and Divergence | 515